System

Abstract

To provide a system.SOLUTION: A system including AI specialized in sales analysis after implementation of Internet advertisement includes: means for learning the transition of sales after implementation of Internet advertisement by AI and outputting the inference results of the transition of sales on the basis of an implementation plan of an advertisement measure; means for presenting the inference results to a first user trying to implement the advertisement measure; and means for analyzing the feeling state of a second user who has browsed an Internet advertisement and optimizing the advertisement measure.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The technology of the present disclosure relates to a system. [Background technology]

[0002] Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to a description of the chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] In analyzing sales after the implementation of internet advertising, it is difficult to accurately predict sales trends after the implementation of the campaign and present the results to users. [Means for solving the problem]

[0005] The present invention uses AI specialized in analyzing sales after the implementation of internet advertising. This AI learns sales trends after the implementation of a campaign and outputs predicted sales trends. This predicted result is then presented to the user. The AI ​​also learns sales trends based on past advertising results and predicts sales trends based on the user's advertising campaign implementation plan. This makes it possible to accurately predict sales trends after the campaign is implemented and present the results to the user. [Brief explanation of the drawings]

[0006] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 2 is a sequence diagram showing a flow of processing in the data processing system according to the first embodiment of the first form example. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1 of Embodiment 1. [Figure 13] FIG. 10 is a sequence diagram showing a processing flow of a data processing system in a second embodiment of the second form example. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 of Embodiment Example 2. [Figure 15] FIG. 10 is a sequence diagram showing the flow of processing in a data processing system according to a third embodiment of the third embodiment. [Figure 16]FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 3 of Embodiment 3. [Figure 17] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in the first embodiment of the first form example when an emotion engine is combined. [Figure 18] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1 of Form Example 1 when an emotion engine is combined. [Figure 19] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in the second embodiment of the second form example when an emotion engine is combined. [Figure 20] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 of Form Example 2 when an emotion engine is combined. [Figure 21] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in the third embodiment of the third form example when an emotion engine is combined. [Figure 22] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 3 of Form Example 3 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0007] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0008] First, the terms used in the following description will be explained.

[0009] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, the processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), or a TPU (TENSOR PROCESSING UNIT (registered trademark)).

[0010] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0011] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0012] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0013] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0014] [First embodiment]

[0015] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0016] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0017] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0018] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0019] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0020] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0021] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0022] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0023] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0024] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0025] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0026] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[0027] "Example 1"

[0028] An example of the present invention is a system including AI specialized in analyzing sales after the implementation of internet advertising. This system collects sales data after the implementation of advertising measures, and the AI ​​learns from this data. The AI ​​predicts future sales trends based on past advertising results and sales data. The prediction results are output and presented to the user in a format that is easy for the user to understand.

[0029] "Example 2"

[0030] Specifically, AI uses deep learning models to learn and predict sales trends. Past advertising results and sales data include information such as the type of ad, ad delivery time, delivery area, and advertising costs. Based on this information, AI predicts sales trends after the implementation of advertising measures.

[0031] "Example 3"

[0032] More specifically, when a user inputs an implementation plan for an advertising campaign, the AI ​​predicts sales trends based on this plan. For example, if a user inputs a plan such as "Next Monday, in the Tokyo area, we will distribute Ad A with an advertising budget of 50,000 yen," the AI ​​will predict sales trends based on this information and present the results to the user.

[0033] The processing flow of each embodiment will be described below.

[0034] "Example 1"

[0035] Step 1: After the advertising campaign is implemented, the system collects sales data.

[0036] Step 2: The collected sales data is learned by AI.

[0037] Step 3: The AI ​​predicts future sales trends based on past advertising results and sales data.

[0038] Step 4: The prediction results are output in a user-friendly format and presented to the user.

[0039] "Example 2"

[0040] Step 1: The AI ​​uses a deep learning model to learn sales trends.

[0041] Step 2: Past advertising results and sales data includes information such as the type of ad, ad delivery time, delivery area, and advertising costs.

[0042] Step 3: Based on this information, the AI ​​predicts sales trends after the advertising campaign is implemented.

[0043] "Example 3"

[0044] Step 1: The user enters an implementation plan for their advertising campaign.

[0045] Step 2: The AI ​​predicts sales trends based on this plan.

[0046] Step 3: For example, if a user inputs a plan such as "Next Monday, in the Tokyo area, we will distribute Ad A with an advertising budget of 50,000 yen," the AI ​​will use this information to predict sales trends.

[0047] Step 4: The prediction results are presented to the user.

[0048] Example 1

[0049] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0050] Quantitatively evaluating the effectiveness of internet advertising and accurately predicting future sales trends is important for maximizing the effectiveness of advertising measures. However, with conventional methods, it was difficult to efficiently collect and analyze sales data after the implementation of advertising measures and predict future sales. Another challenge was presenting the prediction results in a format that was easy for users to understand.

[0051] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0052] In this invention, the server includes means for collecting sales data after the implementation of internet advertising, means for preprocessing the collected sales data, and means for training an artificial intelligence model using the preprocessed data, thereby making it possible to quantitatively evaluate the effectiveness of advertising measures and accurately predict future sales trends.

[0053] "Internet advertising" refers to advertising displayed on the Internet and delivered to users through websites, social media, search engines, etc.

[0054] "Sales data" refers to data relating to the sale of goods and services, and includes information such as the sales amount, sales quantity, and sales date and time.

[0055] "Artificial intelligence" refers to technologies that allow computer systems to learn, reason, and self-correct by imitating human intelligence, and includes machine learning and deep learning.

[0056] "Data collection means" refers to the methods and technologies for obtaining sales data after the implementation of internet advertising, including data acquisition through APIs.

[0057] "Data preprocessing means" refers to methods and techniques for preparing collected sales data in an analyzable format, including filling in missing values, removing outliers, and normalizing data.

[0058] An "artificial intelligence model" is a mathematical model that learns from data and makes predictions and classifications, and includes neural networks and regression models.

[0059] "Learning tools" refers to methods and techniques for training artificial intelligence models using pre-processed data, including machine learning algorithms and deep learning frameworks.

[0060] "Prediction means" refers to methods and technologies for predicting future sales trends using a trained artificial intelligence model, and includes the model's inference function.

[0061] "Presentation means" refers to the method or technique for visually displaying the prediction results to the user, including graphical and tabular presentations.

[0062] The present invention is a system including artificial intelligence specialized for analyzing sales after the execution of internet advertising. Specific embodiments of this system will be described below.

[0063] The server collects sales data after the execution of the internet advertisement. Specifically, the sales data is obtained through an API. For example, the sales data can be obtained using a web service API. The server uses the Python Requests library to send a request to the API endpoint and obtain the sales data.

[0064] Next, the server preprocesses the collected sales data, specifically by imputing missing values, removing outliers, normalizing the data, etc. The server uses the Pandas library to create a data frame, impute missing values, remove outliers, and normalize the data.

[0065] The server then uses the preprocessed data to train an artificial intelligence model. Specifically, it uses TensorFlow (registered trademark) and PyTorch to build a model based on past advertising results and sales data. The server then uses TensorFlow to build a neural network model and trains it on the data.

[0066] After the learning is complete, the server uses the learned artificial intelligence model to predict future sales trends. Specifically, it predicts sales for the next three months. The server uses the trained model to predict sales for the next three months.

[0067] Finally, the terminal presents the prediction results received from the server to the user in the form of graphs and tables. The terminal uses the Matplotlib library to graph the prediction results and display them to the user.

[0068] As a concrete example, the following prompt sentence can be input to a generative AI model:

[0069] Example prompt sentence:

[0070] Collect sales data after implementing advertising campaign A and train the AI ​​model. Based on the advertising campaign and sales data from the past year, predict sales trends for the next three months.

[0071] In this way, users can quantitatively evaluate the effectiveness of their internet advertising and predict future sales. This system enables data-driven decision-making to maximize the effectiveness of advertising initiatives.

[0072] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0073] Step 1:

[0074] The server collects sales data after the implementation of online advertising. As input, it uses an API endpoint for obtaining sales data after the implementation of advertising campaigns. Specifically, the server uses the Python Requests library to send a request to the API and obtains the sales data in JSON format. The output is the obtained sales data.

[0075] Step 2:

[0076] The server preprocesses the collected sales data. It uses the sales data obtained in step 1 as input. Specifically, the server creates a data frame using the Pandas library, imputes missing values, removes outliers, and normalizes the data. The output is the preprocessed sales data.

[0077] Step 3:

[0078] The server uses the preprocessed data to train an artificial intelligence model. The sales data preprocessed in step 2 is used as input. Specifically, the server uses TensorFlow to build a neural network model and trains it on the data. The output is a trained artificial intelligence model.

[0079] Step 4:

[0080] The server uses the trained artificial intelligence model to predict future sales trends. As input, it uses the trained artificial intelligence model and data for the prediction period. Specifically, the server uses the trained model to predict sales for the next three months. The output is the predicted sales data.

[0081] Step 5:

[0082] The terminal presents the forecast results received from the server to the user. As input, it uses the sales data predicted in step 4. Specifically, the terminal uses the Matplotlib library to graph the forecast results and display them to the user. The output is the forecast results presented to the user.

[0083] (Application example 1)

[0084] Next, a description will be given of Application Example 1 of Embodiment Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0085] Accurately analyzing the effectiveness of internet advertising and predicting future sales trends are important challenges for advertisers. However, conventional methods have made it difficult to analyze advertising campaign data in real time and forecast sales. There has also been a lack of ways to present the forecast results in a format that is easy for users to understand. This has made it difficult to make appropriate decisions to maximize the effectiveness of advertising measures.

[0086] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0087] In this invention, the server is a system including an AI specialized in analyzing sales after the implementation of internet advertising. The AI ​​includes means for learning sales trends after the implementation of a campaign and outputting predicted results of sales trends, means for presenting the predicted results to users, means for inputting advertising campaign data and making sales forecasts in real time, and means for displaying the predicted results in graphs and charts. This makes it possible to analyze advertising campaign data in real time and make sales forecasts. In addition, visually displaying the predicted results makes them easy for users to understand and supports appropriate decision-making.

[0088] "Internet advertising" refers to advertising displayed on the Internet, a means of providing information to users through websites and applications.

[0089] "Sales analysis" is the process of collecting and analyzing sales data over a specific period of time to identify sales trends and factors.

[0090] "AI" is an abbreviation for artificial intelligence, a technology that allows computers to imitate human intelligence to learn and reason.

[0091] "Sales trends after the implementation of the campaign" refers to the fluctuations in sales over a certain period of time after the implementation of the advertising campaign.

[0092] The "predicted results" are future sales forecasts based on the data learned by the AI.

[0093] "User" refers to any individual or legal entity that uses the System.

[0094] An "advertising campaign" is a planned advertising activity with a specific purpose, and may include multiple advertising initiatives.

[0095] "Real-time" refers to data being processed and analyzed as it is generated.

[0096] "Graphs and charts" are diagrams that visually represent numerical data, and include bar graphs, line graphs, pie charts, etc.

[0097] A "generative AI model" is an artificial intelligence model trained to perform a specific task, such as making predictions or classifications based on data.

[0098] A "prompt sentence" is a sentence that gives instructions or asks questions to the user, and serves to guide the operation of the system.

[0099] As an embodiment of the present invention, an advertising effectiveness analysis assistant system will be specifically described.

[0100] System configuration

[0101] This system includes AI specialized in analyzing sales after the execution of internet advertising. The system consists of the following main components:

[0102] 1. Data collection module: collects advertising campaign data and sends it to the server.

[0103] 2. Data Preprocessing Module: Cleans the collected data and converts it into a suitable format for analysis.

[0104] 3. AI model: Learns from past advertising results and sales data and predicts future sales trends.

[0105] 4. Prediction result display module: Visually displays the prediction results in graphs and charts.

[0106] Hardware and software used

[0107] Hardware: Smartphones, servers

[0108] Software: Python, pandas, scikit-learn, matplotlib

[0109] Data processing and calculation

[0110] The server processes the data in the following steps:

[0111] 1. Data Loading: The data collection module sends the advertising campaign data to the server as a CSV file, which the server loads.

[0112] 2. Data Preprocessing: The data preprocessing module removes missing values ​​and cleans the data.

[0113] 3. Model training: The AI ​​model trains a linear regression model targeting sales using ad spend, clicks, and impressions as features.

[0114] 4. Sales forecasting: The AI ​​model predicts sales based on new data.

[0115] 5. Displaying the results: The prediction result display module displays the prediction results as graphs and charts.

[0116] Specific examples

[0117] Users prepare their ad campaign data as a CSV file and upload it to the app. Then, they input the data for their new ad campaign (advertising costs, clicks, impressions). The server uses this data to generate sales forecasts and displays the results in graphs.

[0118] Prompt Sentence Examples

[0119] "Upload your ad campaign data and enter new campaign data. The AI ​​will generate sales forecasts and display the results in graphs."

[0120] In this way, the advertising effectiveness analysis assistant system can analyze advertising campaign data in real time and make sales forecasts. In addition, by visually displaying the forecast results, users can easily understand them and help them make appropriate decisions.

[0121] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0122] Step 1:

[0123] The user prepares advertising campaign data as a CSV file and uploads it to the app from their device. The input is the advertising campaign data, and the output is the CSV file sent to the server. Specifically, the user clicks the "Upload Data" button in the app and selects the CSV file from the file selection dialog.

[0124] Step 2:

[0125] The server reads the uploaded CSV file. The input is a CSV file, and the output is the advertising campaign data in a data frame. Specifically, the server converts the CSV file into a data frame using the Python pandas library.

[0126] Step 3:

[0127] The server uses a data preprocessing module to clean the data. The input is the advertising campaign data in a data frame format, and the output is the clean data with missing values ​​removed. Specifically, the server removes missing values ​​and converts the data into the required data type.

[0128] Step 4:

[0129] The server uses the AI ​​model to learn from past advertising results and sales data. The input is clean data, and the output is a trained AI model. Specifically, the server trains a linear regression model using the scikit-learn library.

[0130] Step 5:

[0131] The user inputs data for a new advertising campaign (advertising cost, clicks, impressions) into the app from their device. The input is the new advertising campaign data, and the output is the new data sent to the server. Specifically, the user enters the data into the app's input form and clicks the "Start prediction" button.

[0132] Step 6:

[0133] The server uses the trained AI model to make sales predictions based on new data. The input is new advertising campaign data, and the output is the sales prediction results. Specifically, the server inputs new data into the AI ​​model and obtains the prediction results.

[0134] Step 7:

[0135] The server uses the forecast result display module to visually display the forecast results in graphs and charts. The input is the sales forecast results, and the output is the graphs and charts presented to the user. Specifically, the server uses the matplotlib library to plot the forecast results on a graph and display it on the user's terminal.

[0136] The above are the specific processing steps of the advertising effectiveness analysis assistant system.

[0137] Example 2

[0138] Next, a description will be given of Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0139] Conventional internet advertising effectiveness analysis systems have difficulty accurately predicting sales trends after implementing advertising measures. In addition, there is a lack of effective ways to utilize past advertising results, and highly accurate predictions are required when planning advertising measures.

[0140] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0141] In this invention, the server includes means for collecting past advertising implementation results and sales data, means for preprocessing the collected data, means for training a deep learning model using the preprocessed data, means for a user to input information on new advertising measures, means for predicting sales trends based on the input information on the advertising measures, and means for presenting the prediction results to the user. This makes it possible to predict sales trends after the implementation of advertising measures with high accuracy and provide useful information to the user.

[0142] "Past advertising results" refers to detailed data when advertising was carried out, and includes information such as the type of advertisement, distribution time, distribution area, and advertising costs.

[0143] "Sales data" refers to data that indicates the sales performance of a product during a specific period, and includes information such as sales amount and sales quantity.

[0144] "Preprocessing" refers to the process of converting collected data into a format suitable for analysis and model training, and includes imputing missing values, normalizing data, encoding categorical data, etc.

[0145] A "deep learning model" is an algorithm that uses a multi-layer neural network to learn the characteristics of data and perform predictions and classifications.

[0146] "User" refers to an individual or corporation that uses the system to predict the effectiveness of advertising measures.

[0147] "Information about advertising measures" is detailed data about a new advertisement planned by the user, and includes information such as the type of advertisement, distribution time, distribution area, and advertising cost.

[0148] "Sales trend forecasting" refers to predicting fluctuations in sales after the implementation of advertising campaigns, and is done using a deep learning model.

[0149] The "prediction result" is a predicted value of sales trends calculated by a deep learning model and is presented to the user.

[0150] This invention is a system that predicts sales trends after implementing advertising measures based on past advertising results and sales data. The system is composed of three main elements: a server, a terminal, and a user.

[0151] server

[0152] The server runs on the following hardware and software:

[0153] Hardware: A server machine equipped with a high-performance CPU, GPU, memory, and storage

[0154] Software: Deep learning frameworks such as TensorFlow and PyTorch, database management systems (e.g., MySQL (registered trademark), PostgreSQL)

[0155] The server first collects past advertising results and sales data, which involves retrieving data from the company's database or external APIs, and then stores the collected data in a database.

[0156] The server then preprocesses the collected data, specifically by imputing missing values, normalizing the data, encoding categorical data, etc. The preprocessed data is then converted into a format suitable for input into a deep learning model.

[0157] The server uses the preprocessed data to train a deep learning model. Training is performed using TensorFlow and PyTorch, and the model parameters are optimized. Once trained, the model is used to predict sales trends after the implementation of advertising campaigns.

[0158] Terminal

[0159] The terminal provides an interface for users to input information for new advertising campaigns. The terminal operates using the following hardware and software:

[0160] Hardware: PCs, tablets, smartphones, etc.

[0161] Software: Web browser, dedicated application

[0162] The user inputs information about the new advertising campaign into an input form on the device, including the type of advertisement, distribution time, distribution area, and advertising cost.

[0163] User

[0164] Users are individuals or corporations who use the system to predict the effectiveness of advertising campaigns. Users input information about new advertising campaigns through their terminals and receive prediction results.

[0165] Specific examples

[0166] For example, if a user inputs "A new TV ad will be broadcast in Tokyo from 8:00 PM to 10:00 PM, with an advertising cost of 1 million yen," the server inputs this information into the deep learning model and calculates the predicted sales trend. The server sends the calculation results to the terminal and displays to the user, "Predicted sales trend: 10% increase after one week, 5% increase after one month."

[0167] Prompt Sentence Examples

[0168] "Please predict the sales trend if a new TV ad is broadcast in Tokyo from 8:00 PM to 10:00 PM and the advertising cost is 1 million yen."

[0169] In this way, the server uses a deep learning model to predict the effectiveness of advertising campaigns and provide them to the user.

[0170] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0171] Step 1: Data collection

[0172] The server collects past advertising results and sales data from the company's database or external API. As input, it obtains information such as the type of ad, delivery time, delivery area, and advertising cost. As output, it stores this data in the database. Specifically, the server sends an API request and stores the obtained data in the database.

[0173] Step 2: Data Preprocessing

[0174] The server preprocesses the collected data. As input, it uses the advertising performance results and sales data stored in the database. As output, it generates the preprocessed data. Specifically, the server performs missing value imputation, data normalization, and categorical data encoding. For example, if there are missing values, it imputes them with the mean value, and standardizes numerical data.

[0175] Step 3: Model training

[0176] The server trains a deep learning model using the preprocessed data. It uses the preprocessed data as input and generates a trained deep learning model as output. Specifically, the server uses TensorFlow or PyTorch to build a neural network and optimizes the model parameters using the training data. For training, it uses the Adam optimizer or cross-entropy loss function.

[0177] Step 4: Enter a new advertising strategy

[0178] The user inputs information about a new advertising campaign into an input form on the device. As input, the user provides information such as the type of advertisement, distribution time, distribution area, and advertising cost. As output, the entered information about the advertising campaign is sent to the server. Specifically, the user inputs the information using a web browser or a dedicated application and clicks the send button.

[0179] Step 5: Prediction execution

[0180] The server inputs the information of new advertising campaigns entered by the user into the deep learning model and performs predictions. As input, it uses the information of advertising campaigns provided by the user. As output, it generates predicted sales trends. Specifically, the server inputs the information of advertising campaigns into the deep learning model, and the model calculates sales trends.

[0181] Step 6: View the results

[0182] The server sends the prediction results to the terminal and displays them to the user. The predicted sales trend is used as input. The prediction results are displayed to the user as output. Specifically, the server sends the prediction results to the terminal, and the terminal displays the results on the screen. For example, it displays results such as "Predicted sales trend: 10% increase after one week, 5% increase after one month."

[0183] (Application example 2)

[0184] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0185] Conventional internet advertising effectiveness analysis systems have difficulty accurately predicting sales trends after implementing advertising measures, and do not provide sufficient support for advertisers to develop optimal advertising strategies. Furthermore, there is no way for advertisers to easily predict the effectiveness of their advertising campaigns using smart devices. This has made it difficult for advertisers to optimally allocate advertising costs and implement effective advertising measures.

[0186] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0187] In this invention, the server includes means including an AI specialized in analyzing sales after the implementation of internet advertising, means for the AI ​​to learn sales trends after the implementation of a campaign and output predicted sales trends, means for presenting the predicted results to a user, and means for a user to predict the effectiveness of an advertising campaign using a smart device. This enables advertisers to learn sales trends based on past advertising results and predict sales trends based on information such as the type of advertisement, distribution time, distribution area, and advertising costs. In addition, advertisers can input data for a new advertising campaign using a smart device and predict sales, thereby receiving support for formulating an optimal advertising strategy.

[0188] "Internet advertising" refers to advertising distributed over the Internet, displayed to users through websites, social media, search engines, etc.

[0189] "Sales analysis" is the process of analyzing sales data obtained after implementing advertising measures and evaluating their effectiveness.

[0190] "AI" stands for artificial intelligence, which is a computer system designed to perform specific tasks automatically.

[0191] "Sales trends after the implementation of the campaign" refers to the fluctuations in sales over a certain period of time after the implementation of the advertising campaign.

[0192] The "predicted results" are future sales prospects predicted by AI based on past data.

[0193] "Users" refer to advertisers, marketers, and other people who use this system to predict advertising effectiveness.

[0194] A "smart device" is a portable electronic device with internet connectivity, such as a smartphone or tablet.

[0195] An "advertising campaign" is a series of advertising activities designed to achieve a specific objective.

[0196] "Ad type" refers to the format or media type of the ad, such as banner ad, video ad, or text ad.

[0197] "Serving time" refers to the time period during which an advertisement is displayed to a user.

[0198] "Distribution Area" means the geographic area in which an advertisement is displayed.

[0199] "Advertising costs" are expenses necessary to carry out an advertising campaign.

[0200] "Learning sales trends" is the process of building a model to predict future sales fluctuations based on past sales data.

[0201] "Predicting sales trends" means estimating future sales fluctuations based on past data.

[0202] An "advertising implementation plan" is a plan that includes the specific implementation details and schedule of an advertising campaign.

[0203] A system for carrying out the present invention includes an AI specialized for analyzing sales after the execution of internet advertising. A specific embodiment of this system will be described below.

[0204] System configuration

[0205] The system consists of the following main components:

[0206] 1. Server: Responsible for processing data and running AI models.

[0207] 2. Smart devices: Used by users to predict the effectiveness of advertising campaigns.

[0208] 3. Database: Stores past advertising results and sales data.

[0209] Hardware and software used

[0210] Hardware: Servers, smartphones, tablets

[0211] Software: Python, Pandas, Scikit-learn, TensorFlow / Keras

[0212] Data processing and calculation

[0213] The server processes the data in the following steps to predict sales trends.

[0214] 1. Data loading: The server loads past advertising results and sales data from the database.

[0215] 2. Data preprocessing: Using Pandas, categorical data such as ad type, delivery time, delivery area, and ad cost are converted into dummy variables, and the data is standardized using Scikit-learn.

[0216] 3. Building and training the model: Build a deep learning model using TensorFlow / Keras and train it with the training data.

[0217] 4. Sales forecasting: Input data for a new advertising campaign and use the AI ​​model to predict sales.

[0218] Specific examples

[0219] Forecast sales if an advertiser distributes a new banner ad in Tokyo at 12:00 and spends 50,000 yen on the ad.

[0220] Prompt Sentence Examples

[0221] What is the sales forecast for a new banner ad distributed in Tokyo at 12:00 with an advertising cost of 50,000 yen?

[0222] Processing flow

[0223] 1. User input: The user uses a smart device to input data for a new advertising campaign (ad type, delivery time, delivery area, advertising cost).

[0224] 2. Data transmission: Data is sent from the smart device to the server.

[0225] 3. Sales forecast: The server runs an AI model based on the received data to predict sales trends.

[0226] 4. Presenting the results: The server sends the prediction results to the smart device and presents them to the user.

[0227] In this way, advertisers can predict future sales based on past data and receive support in formulating optimal advertising strategies.

[0228] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0229] Step 1:

[0230] User Input

[0231] The user uses a smart device to enter data for a new advertising campaign (ad type, distribution time, distribution area, advertising cost), which is then formatted by the smart device application and prepared for transmission to the server.

[0232] Input: Ad type, delivery time, delivery area, ad cost

[0233] Output: Formatted ad campaign data

[0234] Step 2:

[0235] Sending data

[0236] The smart device sends the advertising campaign data entered by the user to a server, where it is securely transmitted over the internet.

[0237] Input: Formatted ad campaign data

[0238] Output: Ad campaign data sent to the server

[0239] Step 3:

[0240] Loading data

[0241] The server reads past advertising results and sales data from the database, including information on the type of ad, distribution time, distribution area, advertising cost, and sales.

[0242] Input: Past advertising results and sales data from the database

[0243] Output: Loaded past advertising results and sales data

[0244] Step 4:

[0245] Data Preprocessing

[0246] The server uses Pandas to convert categorical data such as ad type, delivery time, delivery area, and ad cost into dummy variables, and then uses Scikit-learn to standardize the data, preparing it in a format suitable for the AI ​​model.

[0247] Input: Loaded past advertising results and sales data

[0248] Output: Preprocessed data (dummy variables and standardized data)

[0249] Step 5:

[0250] Building and training the model

[0251] The server builds a deep learning model using TensorFlow / Keras and trains the model using the preprocessed data. The trained model has the ability to predict future sales based on past data.

[0252] Input: Preprocessed data

[0253] Output: A trained deep learning model

[0254] Step 6:

[0255] Sales forecast

[0256] The server preprocesses new ad campaign data submitted by users and uses the trained deep learning model to predict sales, outputting the results as a numerical value indicating the likelihood of future sales.

[0257] Input: New advertising campaign data, trained deep learning model

[0258] Output: Sales forecast results

[0259] Step 7:

[0260] Presentation of results

[0261] The server sends the prediction results to the smart device and displays them to the user, who can then check the results on the screen and use them to plan their advertising strategies.

[0262] Input: Sales Forecast Results

[0263] Output: Prediction results displayed on a smart device

[0264] In this way, advertisers can predict future sales based on past data and receive support in formulating optimal advertising strategies.

[0265] Example 3

[0266] Next, a description will be given of a third embodiment of the third embodiment. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0267] With conventional sales analysis systems after the implementation of internet advertising, it was difficult to predict sales trends based on the implementation plan of advertising measures. In addition, there was a lack of means to effectively utilize past advertising results to predict sales trends, making it difficult to evaluate the effectiveness of advertising measures in advance. This made it difficult to optimally allocate advertising expenses and plan effective advertising measures.

[0268] The identification process by the identification processing unit 290 of the data processing device 12 in Example 3 is realized by the following means. In this invention, the server includes a means for a user to input an implementation plan for an advertising measure, a means for a terminal to transmit the input plan to the server, a means for the server to input plan information to the generative AI model, a means for the generative AI model to predict sales trends, a means for the server to receive and transmit the prediction results to the terminal, and a means for the terminal to present the prediction results to the user. This enables the user to predict sales trends based on the implementation plan for the advertising measure and effectively utilize past advertising implementation results.

[0269] A "user" is an entity that uses the system to input an implementation plan for advertising measures and receives forecast results for sales trends.

[0270] A "terminal" is a device through which a user inputs an implementation plan for an advertising campaign and communicates with a server to receive predicted results.

[0271] The "server" is a computer system that receives planning information sent by the user, inputs it into the generative AI model, generates prediction results, and sends them to the terminal.

[0272] A "generative AI model" is an artificial intelligence model that predicts sales trends based on an advertising implementation plan, and has the ability to learn from past advertising implementation results.

[0273] The "advertising measure implementation plan" is planning information including the details and conditions of the advertisement implementation input by the user.

[0274] "Sales trends" is data indicating predicted fluctuations in sales after the implementation of advertising measures.

[0275] "Prediction results" are sales trend forecast data calculated by the generative AI model based on the implementation plan of the advertising campaign.

[0276] This invention is a system that allows a user to input an implementation plan for an advertising campaign and predicts sales trends based on the plan. The system includes a user, a terminal, a server, and a generative AI model.

[0277] The user inputs an implementation plan for advertising measures through the device interface. For example, the user might input a plan such as "Next Monday, in the Tokyo area, we will distribute Ad A with an advertising budget of 50,000 yen." The device receives this input and sends the plan information to the server.

[0278] The server inputs the received plan information into a generative AI model. An example of a generative AI model is an advanced artificial intelligence model such as OpenAI's (registered trademark) GPT-4 (registered trademark). The server converts the plan information into a prompt sentence and passes it to the generative AI model. An example of a prompt sentence is, "The user has input a plan to distribute Ad A in the Tokyo area next Monday with an advertising budget of 50,000 yen. Please predict sales trends based on this plan."

[0279] The generative AI model predicts sales trends based on prompt statements. The model references past data and the results of similar advertising campaigns to generate predictions. For example, it might generate a prediction such as "Monday: 100,000 yen, Tuesday: 80,000 yen, Wednesday: 70,000 yen, Thursday: 60,000 yen, Friday: 50,000 yen."

[0280] The server receives the prediction results obtained from the generative AI model and sends them to the device, which then presents the results to the user, who can then view them through the device interface.

[0281] This system allows users to predict sales trends based on an advertising implementation plan and effectively utilize past advertising implementation results. This makes it easier to optimally allocate advertising expenses and plan effective advertising measures. The flow of the identification process in Example 3 will be described with reference to FIG. 15.

[0282] Step 1:

[0283] The user inputs an implementation plan for the advertising campaign. The user inputs a plan through the device interface, such as "Next Monday, in the Tokyo area, we will distribute Ad A with an advertising budget of 50,000 yen." The input plan information is temporarily stored in the device's memory.

[0284] Step 2:

[0285] The terminal sends the entered plan to the server. The terminal converts the plan information entered by the user into JSON format and sends it to the server using an HTTP request. If the transmission is successful, the terminal displays the message "Plan information sent." The input is the user's plan information, and the output is a transmission success message to the server.

[0286] Step 3:

[0287] The server inputs the plan information into the generative AI model. The server converts the received plan information into a prompt sentence. For example, it generates a prompt sentence such as, "The user has input a plan to distribute Ad A in the Tokyo area next Monday with an advertising budget of 50,000 yen. Please predict sales trends based on this plan." The input is the plan information, and the output is a prompt sentence.

[0288] Step 4:

[0289] The generative AI model predicts sales trends. The generative AI model predicts sales trends based on a prompt statement. The model references data on past advertising campaigns from an internal database to generate a predicted result. For example, the model might generate a predicted result such as "Monday: 100,000 yen, Tuesday: 80,000 yen, Wednesday: 70,000 yen, Thursday: 60,000 yen, Friday: 50,000 yen." The input is the prompt statement, and the output is the predicted sales trend.

[0290] Step 5:

[0291] The server receives the prediction results and sends them to the device. The server receives the prediction results obtained from the generative AI model, converts them to JSON format, and sends them to the device. If the transmission is successful, the server records a log stating "Prediction results sent." The input is the prediction results, and the output is a successful transmission message to the device.

[0292] Step 6:

[0293] The device presents the prediction results to the user. The device displays the prediction results received from the server on the screen. The user can check the prediction results on the device screen, such as "Monday: 100,000 yen, Tuesday: 80,000 yen, Wednesday: 70,000 yen, Thursday: 60,000 yen, Friday: 50,000 yen." The input is the prediction results, and the output is what is displayed to the user.

[0294] (Application example 3)

[0295] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0296] Conventional sales analysis systems for post-implementation internet advertising have had the problem of making it difficult to predict the effectiveness of advertising measures in advance, preventing advertisers from formulating optimal advertising strategies. Furthermore, there was also the problem of insufficient learning of sales trends based on past advertising results, resulting in low prediction accuracy.

[0297] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 3 is realized by the following means. In this invention, the server includes a means for a user to input an implementation plan for an advertising measure, a means including a generative AI model that predicts sales trends based on the implementation plan for the advertising measure, and a means for presenting the prediction results to the user. This allows advertisers to simulate the effects of their advertising measures in advance and develop optimal advertising strategies. In addition, the inclusion of a means for learning sales trends based on past advertising implementation results can improve prediction accuracy. Furthermore, the inclusion of an application to be installed on a smartphone allows users to easily check the predicted results of their advertising measures.

[0298] A "user" is an entity that inputs an implementation plan for advertising measures and checks the predicted results.

[0299] "Advertising implementation plan" refers to the specific details of the plan, such as the date, area, budget, and type of advertisement to be implemented.

[0300] A "generative AI model" is an artificial intelligence model that predicts sales trends based on the implementation plan of advertising campaigns entered by the user.

[0301] "Sales trends" is data showing fluctuations in sales after the implementation of advertising measures.

[0302] The "prediction result" is the forecast of future sales trends calculated by the generative AI model based on the implementation plan of the advertising campaign.

[0303] "Past advertising results" refers to the details of advertising measures previously implemented and the sales data obtained as a result.

[0304] A "smartphone" is a type of mobile phone, a multi-function device that can connect to the Internet and install applications.

[0305] An "application" is a software program that is installed on a smartphone and provides a specific function.

[0306] The system for implementing this invention uses a generative AI model that predicts sales trends based on an advertising implementation plan entered by the user. The system can be operated through an application installed on the user's smartphone.

[0307] The server receives the advertising implementation plan entered by the user and predicts sales trends based on that plan. Specifically, the server uses a generative AI model that learns sales trends based on past advertising results. This generative AI model is built using Python and its libraries (numpy, scikit-learn). The server calculates the prediction results and presents them to the user.

[0308] Users use an application installed on their smartphone to input plans for advertising initiatives. For example, they might input a plan such as "Next Monday, in the Tokyo area, we will distribute Ad A with an advertising budget of 50,000 yen." This input is done in text format.

[0309] An example of a prompt statement is as follows:

[0310] Please enter your advertising implementation plan.

[0311] Example: Next Monday, Ad A will be distributed in the Tokyo area with an advertising budget of 50,000 yen.

[0312] The server receives this input and uses the generative AI model to predict sales trends. The prediction results are displayed to users in real time via a smartphone application. This allows users to simulate the effects of their advertising campaigns in advance and develop optimal advertising strategies.

[0313] For example, if a user inputs a plan to "distribute Ad A next Monday in the Tokyo area with an advertising budget of 50,000 yen," the server will predict sales trends based on past data and present the results to the user as "Predicted sales: 15,000 yen." In this way, users can check the effectiveness of their advertising measures in advance and make appropriate decisions.

[0314] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[0315] Step 1:

[0316] The user launches the smartphone application and inputs the implementation plan for the advertising campaign, including the date, region, budget, and type of advertisement. The input data is sent to the server in text format.

[0317] Step 2:

[0318] The server receives the advertising plan sent by the user, analyzes the received data, and extracts necessary information (e.g., date, region, budget, and type of ad). Based on the analysis results, it prepares the data to be input into the generative AI model.

[0319] Step 3:

[0320] The server retrieves relevant data from a database of past advertising results. The retrieved data includes details of past advertising campaigns and the resulting sales data. This data is used to build a training dataset for the generative AI model.

[0321] Step 4:

[0322] The server uses a generative AI model to predict sales trends based on the implementation plan for advertising measures entered by the user. Specifically, it inputs new advertising measure data into a model trained on past data and predicts sales trends. The prediction results are calculated and output as numerical data.

[0323] Step 5:

[0324] The server sends the forecast results to the smartphone application to present to the user. The sent data is numerical data of the forecast sales.

[0325] Step 6:

[0326] The user's smartphone displays the received prediction results on the application screen, allowing the user to check the predicted sales results and simulate the effectiveness of advertising measures in advance.

[0327] Specifically, when a user inputs a plan such as "Distributing Ad A next Monday in the Tokyo area with an advertising budget of 50,000 yen," the server predicts sales trends based on past data and presents the results to the user as "Predicted sales: 15,000 yen." In this way, users can check the effectiveness of their advertising measures in advance and make appropriate decisions.

[0328] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0329] "Example 1"

[0330] In one embodiment of the present invention, the emotion engine recognizes the user's emotions. Specifically, it infers emotions from the user's facial expressions, tone of voice, and the language used in text messages. For example, when a user is happy, their voice tone tends to be higher and their language tends to be more positive. The AI ​​analyzes this information to recognize the user's emotional state.

[0331] "Example 2"

[0332] The emotion engine also evaluates the effectiveness of advertising campaigns based on the user's emotional state. Specifically, it analyzes how the user's emotional state changed after the advertising campaign was implemented. For example, if the user's emotional state clearly improved after the advertising campaign was implemented, the advertising campaign is evaluated as effective.

[0333] "Example 3"

[0334] Furthermore, the emotion engine suggests optimizing advertising measures based on the user's emotional state. Specifically, it learns patterns of advertising measures when the user's emotional state is good, and suggests new advertising measures based on those patterns. For example, if a user sees a specific ad A when they are happy, it will determine that ad A is likely to be well-received by the user and suggest a similar advertising measure.

[0335] The processing flow of each embodiment will be described below.

[0336] "Example 1"

[0337] Step 1: The emotion engine infers emotions from the user's facial expressions, tone of voice, and wording in text messages.

[0338] Step 2: AI analyzes this information and recognizes the user's emotional state.

[0339] "Example 2"

[0340] Step 1: After the advertising campaign is implemented, the emotion engine analyzes how the user's emotional state has changed.

[0341] Step 2: If the user's emotional state is clearly improved after the ad campaign is implemented, the ad campaign is evaluated as effective.

[0342] "Example 3"

[0343] Step 1: The emotion engine learns advertising patterns when the user is in a positive emotional state.

[0344] Step 2: Propose new advertising strategies based on those patterns.

[0345] Step 3: If a user sees a specific ad A while they are happy, we determine that ad A is likely to be well-received by the user and suggest a similar advertising strategy.

[0346] Example 1

[0347] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0348] Accurately analyzing the effectiveness of internet advertising and predicting future sales trends are important challenges for advertisers. However, conventional systems often require manual data collection, preprocessing, and output of prediction results, which is inefficient. Furthermore, they are unable to provide feedback that takes user emotions into account, making it difficult to use the data to improve advertising strategies. To solve these challenges, a system is needed that can automatically collect, preprocess, predict, and output data, as well as the ability to recognize and analyze user emotions.

[0349] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0350] In this invention, the server includes means for collecting sales data after the implementation of internet advertising, means for preprocessing the collected sales data, means for training an artificial intelligence model based on the preprocessed data, means for predicting future sales trends using the trained artificial intelligence model, means for outputting the prediction results in a format that is easy for the user to understand, means for recognizing user emotions, and means for analyzing the recognized emotion data. This enables automatic collection, preprocessing, prediction, and output of data, and further enables feedback that takes user emotions into consideration.

[0351] "Internet advertising" refers to advertising displayed on the Internet and delivered to users through websites, social media, search engines, etc.

[0352] "Sales data" refers to information regarding sales obtained after the implementation of advertising measures, including the number of clicks, conversion rate, sales amount, etc.

[0353] "Artificial intelligence" refers to technologies that allow computer systems to learn, reason, and self-correct by imitating human intelligence, and includes machine learning and deep learning.

[0354] "Preprocessing" refers to the process of converting collected data into a format suitable for analysis and learning, and includes filling in missing values, removing outliers, and normalizing data.

[0355] An "artificial intelligence model" is a mathematical model that is built to learn from data and perform specific tasks, including neural networks.

[0356] "Forecasting" means estimating future events and trends based on past data, and includes forecasting sales trends.

[0357] "Output" means that the system presents the processing results to the user, including display in the form of graphs or reports.

[0358] "Emotion recognition" is a technology that estimates a user's emotions from their facial expressions, tone of voice, and the wording of text messages, and it uses cameras, microphones, and natural language processing technology.

[0359] "Emotional data" refers to information about a user's emotional state, including emotional categories such as positive, negative, and neutral.

[0360] "Analysis" involves examining collected data in detail to find meaning and patterns, and includes processing to understand user feedback.

[0361] This invention is a system that includes artificial intelligence specialized in analyzing sales after the implementation of internet advertising. This system predicts future sales trends based on sales data collected after the implementation of advertising campaigns and presents them to users. It also has the ability to recognize user emotions and analyze feedback.

[0362] Hardware and software used

[0363] The server uses the following hardware and software:

[0364] Hardware: A server machine with a powerful processor, sufficient memory, and storage.

[0365] Software: Use APIs for data collection and machine learning frameworks such as TensorFlow and PyTorch for data preprocessing and training artificial intelligence models

[0366] The terminal uses the following hardware and software:

[0367] Hardware: A computer or smart device with a camera and microphone

[0368] Software: Emotion recognition software using natural language processing technology

[0369] Program processing explanation

[0370] The server collects sales data after the internet advertising campaign is executed. Specifically, the data is obtained via API from Google Analytics and advertising platforms (e.g., Google Ads, Facebook Ads). The collected data undergoes preprocessing such as filling in missing values, removing outliers, and normalizing the data.

[0371] Based on the preprocessed data, the server trains an artificial intelligence model using TensorFlow or PyTorch. The trained model is then used to predict future sales trends. The prediction results are presented to the user in the form of a web dashboard or PDF report.

[0372] The device recognizes emotions from the user's facial expressions, tone of voice, and the wording of text messages using a camera, microphone, and natural language processing technology. The recognized emotion data is analyzed on the server and reflected as feedback to the user.

[0373] Specific examples

[0374] For example, consider the case where sales trends for November 2023 are predicted based on data from an online advertising campaign for a new product that was conducted in October 2023. The server collects data such as click counts, conversion rates, and sales amounts from Google Analytics and advertising platforms, and performs preprocessing. A neural network is built using TensorFlow, and a model is trained based on past advertising results and sales data. The trained model is used to predict sales for the next advertising campaign, and the results are displayed on a web dashboard.

[0375] Prompt Sentence Examples

[0376] "Based on data from an online advertising campaign for a new product in October 2023, please predict sales trends for November 2023. Display the results in a graph and analyze the user's emotional state."

[0377] In this way, the system not only analyzes the effectiveness of advertising campaigns and predicts future sales, but also recognizes user sentiment and provides feedback.

[0378] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0379] Step 1: Data collection

[0380] The server collects sales data after the execution of online advertising. Specifically, it obtains data from Google Analytics and advertising platforms (e.g., Google Ads, Facebook Ads) via API.

[0381] Input: Data such as clicks, conversion rate, and sales amount for your ad campaign

[0382] Output: Collected sales data

[0383] What happens: The server periodically sends API requests to retrieve data about advertising campaigns, which is then stored in a database.

[0384] Step 2: Data Preprocessing

[0385] The server preprocesses the collected sales data, specifically by filling in missing values, removing outliers, and normalizing the data.

[0386] Input: Collected sales data

[0387] Output: Preprocessed data

[0388] What it does: The server imputes missing values ​​with the average, removes abnormally high sales data, and standardizes the data to convert it into a format that is easy for the AI ​​model to learn.

[0389] Step 3: Training the AI ​​model

[0390] The server uses the preprocessed data to train an AI model, which uses frameworks such as TensorFlow and PyTorch.

[0391] Input: Preprocessed data

[0392] Output: A trained AI model

[0393] Specific operation: The server uses TensorFlow to build a neural network and trains the model based on past advertising results and sales data.

[0394] Step 4: Sales Forecast

[0395] The server uses the trained AI model to predict future sales trends.

[0396] Input: Trained AI model, current advertising data

[0397] Output: Predicted future sales trends

[0398] What it does: The server predicts sales for the next advertising campaign and generates the results in a graph format.

[0399] Step 5: Output the results

[0400] The server outputs the prediction results in a user-friendly format, such as a web dashboard or PDF report.

[0401] Input: Forecast results of future sales trends

[0402] Output: Web dashboard, PDF report

[0403] What it does: The server displays the prediction results on a web dashboard and allows users to download them as a PDF report.

[0404] Step 6: Emotion Recognition

[0405] The device uses a camera, microphone, and natural language processing technology to recognize emotions from the user's facial expressions, tone of voice, and the wording of text messages.

[0406] Input: User's facial expressions, tone of voice, text messages

[0407] Output: Recognized emotion data

[0408] Specific operation: When a user accesses the system through a webcam, the device uses the camera and microphone to analyze the user's facial expressions and tone of voice to estimate their emotions.

[0409] Step 7: Analyze the sentiment data

[0410] The server analyzes the collected emotional data and recognizes the user's emotional state, which allows it to understand the user's feedback more effectively.

[0411] Input: Recognized emotion data

[0412] Output: Parsed emotional feedback

[0413] What it does: If a user gives positive feedback, the server will recognize that sentiment and incorporate it into the strategy for the next advertising campaign.

[0414] (Application example 1)

[0415] Next, a description will be given of Application Example 1 of Embodiment Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0416] Conventional internet advertising effectiveness analysis systems can predict future sales trends based on sales data after the implementation of advertising measures, but they do not evaluate the effectiveness of advertising measures taking into account the emotional state of the user. This makes it difficult to propose optimal advertising measures to maximize advertising effectiveness.

[0417] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0418] In this invention, the server includes means for learning sales trends after the implementation of internet advertising and outputting predicted results of sales trends, means for presenting the predicted results to the user, means for analyzing the user's facial expressions and tone of voice to recognize the emotional state, and means for evaluating the effectiveness of advertising measures based on the emotional state. This makes it possible to evaluate the effectiveness of advertising measures taking the user's emotional state into consideration, and to propose optimal advertising measures to maximize the effectiveness of advertising.

[0419] "Internet advertising" refers to advertising displayed on the Internet, a means of providing information to users through websites and applications.

[0420] "Sales analysis" is the process of analyzing sales data obtained after implementing advertising measures and evaluating their effectiveness.

[0421] "AI" stands for artificial intelligence, a technology that learns from data and makes predictions and judgments.

[0422] "Sales trends" is data that indicates fluctuations in sales over a certain period of time.

[0423] The "predicted results" are the results of predicting future sales trends based on the data learned by the AI.

[0424] "User" means any individual or entity viewing an Internet advertisement.

[0425] "Facial expressions" refer to the movements of a user's face and changes in their facial muscles, and provide clues to their emotional state.

[0426] "Tone of voice" refers to vocal characteristics such as the pitch, intensity, and intonation of a user's voice, which provide clues to their emotional state.

[0427] "Emotional state" refers to the user's psychological state, including emotions such as joy, sadness, and anger.

[0428] An "advertising program" is an advertising implementation method or strategy designed to achieve a specific objective.

[0429] "Effectiveness evaluation" is the process of evaluating the effectiveness of advertising measures.

[0430] The system for carrying out the present invention performs sales analysis taking into account the emotional state of the user in order to maximize the effectiveness of internet advertising. Specific embodiments are described below.

[0431] System Configuration

[0432] This system is configured using the following hardware and software.

[0433] Hardware: Smartphone camera, smartphone microphone

[0434] Software: OpenCV, TensorFlow / Keras, librosa

[0435] Program processing

[0436] 1. Facial Expression Recognition

[0437] The device (smartphone) captures the user's facial expressions using a camera, performs face detection using OpenCV, converts the detected facial region to grayscale, and then inputs it into an emotion recognition model, allowing the user's emotional state to be recognized in real time.

[0438] 2. Voice Emotion Recognition

[0439] The device records the user's voice using a microphone. Using librosa, the voice data is converted into MFCC features and input into a voice emotion recognition model. This allows the device to recognize the user's emotional state from the tone of their voice.

[0440] 3. Sales forecasting

[0441] The server collects and standardizes advertising campaign data, then inputs it into a sales forecasting model, which allows future sales trends to be predicted.

[0442] 4. Effectiveness evaluation

[0443] The server evaluates the effectiveness of advertising measures based on the user's emotional state and sales forecast results, allowing it to propose optimal advertising measures to maximize advertising effectiveness.

[0444] Specific examples

[0445] For example, if a user is viewing an advertisement on their smartphone and the camera captures their facial expressions, and the emotion recognition model detects "joy," the effectiveness of the advertising campaign can be evaluated based on that information, and future sales can be predicted.

[0446] Prompt Sentence Examples

[0447] When a user views an ad on their smartphone, the camera and microphone are used to analyze the user's facial expressions and tone of voice in real time, and an emotion engine is used to recognize the user's emotional state. Furthermore, sales data after viewing the ad is collected, and the AI ​​learns from this to predict future sales trends. This allows the system to propose optimal advertising measures to maximize the effectiveness of the ad.

[0448] In this way, it becomes possible to evaluate the effectiveness of advertising measures taking into account the emotional state of the user, and to propose optimal advertising measures to maximize the effectiveness of advertising.

[0449] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0450] Step 1:

[0451] The device captures the user's facial expressions using the smartphone camera. The input is the camera image, and the output is the captured facial image. Face detection is performed using OpenCV, and the detected facial area is converted to grayscale. This results in image data suitable for facial expression recognition.

[0452] Step 2:

[0453] The device inputs the grayscale facial image into the emotion recognition model. The input is a grayscale image, and the output is the emotion recognition result. The emotion recognition model is executed using TensorFlow / Keras to recognize the user's emotional state (e.g., joy, sadness, anger, etc.). This allows the user's emotional state to be grasped in real time.

[0454] Step 3:

[0455] The device records the user's voice using the smartphone's microphone. The input is audio data, and the output is a recorded audio file. Librosa is used to convert the audio data into MFCC features, which converts the audio data into a format suitable for emotion recognition models.

[0456] Step 4:

[0457] The device inputs the voice data converted into MFCC features into a voice emotion recognition model. The input is the MFCC features, and the output is the voice emotion recognition result. The voice emotion recognition model is executed using TensorFlow / Keras, and the emotional state of the user is recognized from the tone of their voice. This makes it possible to understand the user's emotional state from their voice.

[0458] Step 5:

[0459] The server collects and standardizes advertising campaign data. The input is advertising campaign data, and the output is standardized data. StandardScaler is used to standardize the data and convert it into a format suitable for the sales forecasting model. This reduces data variability and improves model accuracy.

[0460] Step 6:

[0461] The server inputs the standardized advertising campaign data into the sales forecasting model. The input is the standardized data, and the output is the sales forecast result. The sales forecasting model is executed using TensorFlow / Keras to predict future sales trends. This allows the effectiveness of advertising campaigns to be evaluated numerically.

[0462] Step 7:

[0463] The server evaluates the effectiveness of advertising measures based on the user's emotional state and sales forecast results. The inputs are emotion recognition results and sales forecast results, and the output is the results of evaluating the effectiveness of advertising measures. This makes it possible to evaluate the effectiveness of advertising measures taking the user's emotional state into account, and to propose optimal advertising measures.

[0464] Example 2

[0465] Next, a description will be given of Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0466] Conventional internet advertising effectiveness analysis systems not only predict sales trends after the implementation of advertising measures, but also do not evaluate the effectiveness of advertising measures by taking into account the emotional state of the user. This makes it difficult to accurately evaluate the true effectiveness of advertising measures. In addition, data collection and preprocessing for evaluating the effectiveness of advertising measures are often done manually, resulting in inefficiencies.

[0467] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0468] In this invention, the server includes means for collecting past advertising implementation results and sales data, means for saving the collected data in a database, means for preprocessing the saved data, means for training a deep learning model using the preprocessed data, means for a user to input information about a new advertising campaign, means for predicting sales trends based on the input information, means for presenting the prediction results to the user, means for collecting the user's emotional state after the advertising campaign is implemented, means for sending the collected emotional data to an emotion engine, means for receiving analysis results from the emotion engine, means for evaluating the effectiveness of the advertising campaign based on the analysis results from the emotion engine, and means for presenting the evaluation results to the user. This makes it possible not only to predict sales trends after the implementation of advertising campaigns, but also to evaluate the effectiveness of advertising campaigns taking the user's emotional state into consideration.

[0469] "Past advertising results" refers to data on previously conducted advertising campaigns, including information such as the type of advertisement, the time of advertisement distribution, the distribution area, and the cost of advertisements.

[0470] "Sales data" refers to information relating to the sales quantity and sales amount of a product during a specific period.

[0471] A "database" is a system for efficiently storing, managing, and searching collected data.

[0472] "Preprocessing" refers to processes used to convert data into a format suitable for analysis and learning, such as filling in missing values, removing outliers, and normalizing data.

[0473] A "deep learning model" is an algorithm that uses a multi-layer neural network to learn the characteristics of data and perform predictions and classifications.

[0474] "User" means an individual or organization that uses the system to analyze and evaluate the effectiveness of advertising measures.

[0475] An "advertising initiative" is an advertising campaign or promotional activity designed to achieve a specific objective.

[0476] "Emotional state" refers to the emotional changes and psychological reactions that users experience after implementing advertising campaigns.

[0477] An "emotion engine" is software or a system for analyzing a user's emotional state and providing the results.

[0478] The "evaluation result" is the result of quantitatively or qualitatively evaluating the effectiveness of the advertising campaign.

[0479] This invention is a system that predicts and evaluates the effectiveness of advertising measures based on past advertising results and sales data. The system consists of three main elements: a server, a terminal, and a user.

[0480] Data collection and preprocessing

[0481] The server collects past advertising results and sales data. This data includes information such as the type of ad, ad delivery time, delivery area, and advertising cost. The collected data is stored in a MySQL database. The server then performs preprocessing on the data. Specifically, it performs tasks such as filling in missing values, removing outliers, and normalizing the data. For example, it removes data with abnormally high advertising costs and fills in missing sales data with the average value.

[0482] Data training

[0483] The server trains a deep learning model using the preprocessed data. This model is built using TensorFlow. The model architecture includes multiple layers (input layer, hidden layer, and output layer). The server trains the model using the type of ad, distribution time, distribution area, and advertising cost as input data and sales trends as output data. For example, the model can be trained using data from the past year.

[0484] Sales trend forecast

[0485] When a user wants to implement a new advertising campaign, they input the information from their device. For example, they input information about the advertising campaign, such as "broadcast a new TV commercial nationwide with advertising costs of 1 million yen." The server uses a deep learning model based on the input information to predict sales trends. For example, it generates a prediction that sales will increase by 20% in the next month. The prediction result is displayed on the device, and the user can check it.

[0486] Emotional state analysis

[0487] After the advertising campaign is implemented, the terminal conducts a survey of the user to collect the user's emotional state. For example, the user may input an emotional state such as "I felt good after seeing the advertisement." The server then sends the collected emotional data to the IBM Watson (registered trademark) Tone Analyzer. The emotion engine analyzes the user's emotional state and sends the results to the server.

[0488] Evaluating the effectiveness of advertising measures

[0489] The server evaluates the effectiveness of the advertising campaign based on the analysis results of the emotion engine. For example, if the user's emotional state improves, the advertising campaign is evaluated as effective. The evaluation results are displayed on the device so that the user can check them.

[0490] Examples of specific examples and prompts

[0491] For example, if a user wants to implement an advertising campaign such as "broadcasting a new TV commercial nationwide with an advertising budget of 1 million yen," the user inputs this information into the device. Using a deep learning model, the server displays a predicted result that sales will increase by 20% in the next month. Furthermore, if the user responds after implementing the advertising campaign that "watching the ad made me feel good," the emotion engine will evaluate the advertising campaign as effective.

[0492] Example prompt sentence:

[0493] "We would like to implement a new advertising initiative. The ad will be a television commercial, aired during prime time, distributed nationwide, and cost 1 million yen. Please tell us your projected sales trends."

[0494] "Please tell us how the user's emotional state changed after implementing the advertising campaign. If the user's emotional state improved, please rate the advertising campaign as effective."

[0495] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0496] Step 1:

[0497] The server collects past advertising results and sales data. Specifically, it obtains data from marketing systems and sales management systems. As input, it receives information such as the type of advertisement, advertisement distribution time, distribution area, and advertising cost, and as output, it stores this data in a database.

[0498] Step 2:

[0499] The server stores the collected data in a database. Specifically, it stores the data in a MySQL database. It receives the data collected in step 1 as input and generates the data stored in the database as output.

[0500] Step 3:

[0501] The server preprocesses the stored data by imputing missing values, removing outliers, and normalizing the data. It receives raw data stored in the database as input and generates preprocessed data as output. For example, it removes data with abnormally high advertising costs and imputes missing sales data with the average value.

[0502] Step 4:

[0503] The server trains a deep learning model using the preprocessed data. Specifically, it uses TensorFlow to build a model and trains it using the training data as input. It receives the preprocessed data as input and generates a trained deep learning model as output.

[0504] Step 5:

[0505] The user inputs information about a new advertising campaign into the terminal. Specifically, the user inputs information such as the type of advertisement, distribution time, distribution area, and advertising cost. Detailed information about the advertising campaign is received as input, and this information is sent to the server as output.

[0506] Step 6:

[0507] The server predicts sales trends based on the input information. Specifically, it makes the prediction using a trained deep learning model. It receives advertising campaign information sent by the user as input and generates predicted sales trends as output. For example, it generates a prediction that sales will increase by 20% in the next month.

[0508] Step 7:

[0509] The server displays the prediction results on the terminal. Specifically, it presents the predicted sales trends to the user. It receives the prediction results as input and displays them on the terminal screen as output.

[0510] Step 8:

[0511] The user inputs their emotional state into the terminal after the implementation of the advertising campaign. Specifically, the emotional state is input in the form of a questionnaire. The emotional state after the implementation of the advertising campaign is received as input, and this information is sent to the server as output.

[0512] Step 9:

[0513] The server sends the collected emotion data to the emotion engine. Specifically, it sends the data to the emotion engine and requests it to analyze. It receives the emotion data sent by the user as input and sends it to the emotion engine as output.

[0514] Step 10:

[0515] The server receives the analysis results of the emotion engine. Specifically, the server receives the analysis results from the emotion engine. The server receives the analysis results from the emotion engine as input and stores the analysis results as output.

[0516] Step 11:

[0517] The server evaluates the effectiveness of the advertising campaign based on the analysis results of the emotion engine. Specifically, it evaluates whether the user's emotional state has improved. It receives the analysis results of the emotion engine as input and generates an evaluation result of the effectiveness of the advertising campaign as output.

[0518] Step 12:

[0519] The server displays the evaluation results on the terminal. Specifically, it presents the results of the evaluation of the effectiveness of the advertising campaign to the user. It receives the evaluation results as input and displays them on the terminal screen as output.

[0520] (Application example 2)

[0521] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0522] Conventional internet advertising effectiveness analysis systems are specialized in predicting sales trends, but do not take into account the emotional state of users when evaluating the effectiveness of advertising measures, making it difficult to accurately evaluate the true effectiveness of advertising. In addition, there is a lack of means to analyze user feedback text and calculate emotional scores, resulting in a lack of information needed to improve advertising measures. This makes it difficult for advertisers to plan and implement effective advertising measures.

[0523] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0524] In this invention, the server is a system including an AI specialized in analyzing sales after the implementation of internet advertising, and the AI ​​includes means for learning sales trends after the implementation of the campaign and outputting predicted results of the sales trends, means for presenting the predicted results to the user, means for evaluating the effectiveness of the advertising campaign based on the emotional state of the user, and means for analyzing the text of the user's feedback and calculating an emotional score. This makes it possible to simultaneously evaluate sales forecasts for advertising campaigns and the emotional state of the user.

[0525] "Internet advertising" refers to advertising delivered via the Internet and displayed to users through websites, social media, search engines, etc.

[0526] "Sales analysis" is the process of collecting and analyzing sales data over a specific period of time to identify sales trends and factors.

[0527] "AI" stands for artificial intelligence, a technology that allows machines to mimic human intelligence to learn, reason, and self-correct.

[0528] "Sales trends" are data that indicate fluctuations in sales over a certain period of time, and represent increases or decreases in sales over time.

[0529] The "estimated results" are the expected future sales figures that the AI ​​predicts based on past data.

[0530] "User" refers to consumers and advertisers who view internet advertisements.

[0531] "Emotional state" refers to the emotions a user feels when viewing an advertisement, including emotions such as joy, sadness, and excitement.

[0532] "Advertising measures" refer to advertising strategies and means planned and implemented to achieve a specific purpose.

[0533] "Evaluating effectiveness" means measuring and evaluating the extent to which advertising measures have achieved the desired results.

[0534] "Feedback" refers to opinions and feedback provided by users regarding advertisements.

[0535] "Text analysis" is the process of analyzing text data using natural language processing techniques to extract meaning and sentiment.

[0536] An "emotion score" is a numerical representation of a user's emotional state obtained through text analysis.

[0537] As an embodiment of the present invention, a system for analyzing the effectiveness of internet advertising and predicting sales trends is constructed. The system includes a server, a user terminal, and an AI model.

[0538] The server collects past advertising results and sales data, and uses this data to train a deep learning model. Specifically, it uses information such as ad type, distribution time, distribution area, and advertising cost as features to build a model that predicts sales trends. The server implements a random forest regression model using Python, and performs data preprocessing, model training, and predictions.

[0539] The user terminal provides an interface for advertisers to input information about new advertising initiatives. The user inputs information such as the type of advertisement, distribution time, distribution area, and advertising cost, and sends it to the server. The server uses this information to predict sales trends and returns the results to the user terminal.

[0540] Furthermore, the server collects user feedback and performs text analysis. Specifically, it uses the TextBlob library to analyze user feedback and calculate an emotion score. The emotion score is a numerical representation of the user's emotional state and is used to evaluate the effectiveness of advertising campaigns.

[0541] For example, if the type of advertisement is "video," the distribution time is "evening," the distribution area is "Tokyo," and the advertisement cost is "50,000 yen," the user enters the information as follows:

[0542] Ad Type: Video

[0543] Delivery time: Evening

[0544] Distribution area: Tokyo

[0545] Advertising cost: 50,000 yen

[0546] User Feedback: This ad made me so happy and excited!

[0547] The server uses this information to predict sales trends and analyzes user feedback to calculate an emotion score, allowing advertisers to evaluate the effectiveness of their advertising campaigns from both sales and emotion perspectives.

[0548] This system will provide advertisers with valuable information to plan and implement effective advertising strategies.

[0549] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0550] Step 1:

[0551] The server collects past advertising results and sales data. This includes information such as the type of ad, delivery time, delivery area, and advertising cost. This data is stored in a database for later processing. The input is past advertising data and sales data, and the output is the data stored in the database.

[0552] Step 2:

[0553] The server trains a deep learning model based on the collected data. Specifically, it implements a random forest regression model using Python and performs data preprocessing (such as imputing missing values ​​and normalizing the data). Next, it splits the data into training and test data and trains the model. The input is advertising data and sales data obtained from the database, and the output is the trained model.

[0554] Step 3:

[0555] The user terminal provides an interface for the advertiser to input information about new advertising initiatives. The user inputs information such as the type of advertisement, distribution time, distribution area, and advertising cost, and sends it to the server. The input is the advertising initiative information entered by the advertiser, and the output is the data sent to the server.

[0556] Step 4:

[0557] The server predicts sales trends based on the advertising campaign information sent by the user. Using the trained model, it predicts sales based on the features of the advertising campaigns entered. The input is the advertising campaign information sent by the user, and the output is predicted sales data.

[0558] Step 5:

[0559] The server returns the predicted sales data to the user terminal, which then displays the prediction results to the advertiser. The input is the predicted sales data, and the output is the prediction results displayed on the user terminal.

[0560] Step 6:

[0561] The user terminal collects user feedback on advertising campaigns. Users input their impressions and opinions on advertisements in text format and send them to the server. The input is the user feedback, and the output is the feedback data sent to the server.

[0562] Step 7:

[0563] The server analyzes the user's feedback text and calculates the emotion score. Specifically, it uses the TextBlob library to analyze the feedback and quantify the positive and negative emotions. The input is the user's feedback data, and the output is the calculated emotion score.

[0564] Step 8:

[0565] The server evaluates the effectiveness of the advertising campaign based on the calculated emotion score. It combines the sales forecast and emotion score to evaluate the overall effectiveness of the advertising campaign and returns the results to the user's device. The input is the sales forecast data and emotion score, and the output is the evaluation result of the advertising campaign's effectiveness.

[0566] Step 9:

[0567] The user terminal displays the results of the effectiveness evaluation of the advertising campaign to the advertiser. The advertiser uses this information to plan and implement their next advertising campaign. The input is the results of the effectiveness evaluation of the advertising campaign, and the output is the evaluation result displayed to the advertiser.

[0568] Example 3

[0569] Next, a description will be given of a third embodiment of the third embodiment. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0570] Conventional post-sales analysis systems for internet advertising have had issues with accurately predicting the effectiveness of advertising measures and being unable to propose optimal advertising measures that take into account the emotional state of the user. This makes it difficult to find appropriate measures to maximize advertising effectiveness, which can result in wasted advertising costs and sluggish sales.

[0571] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[0572] In this invention, the server includes means for a user to input an implementation plan for an advertising campaign, means for a terminal to send the input information to the server, means for the server to preprocess the received information, means for the server to input a prompt sentence to a generative AI model and predict sales trends, means for the server to present the prediction results to the user, means for the server to collect the user's emotional state, and means for the server to propose optimization of the advertising campaign using an emotion engine. This makes it possible to accurately predict the effectiveness of advertising campaigns and further propose optimal advertising campaigns that take the user's emotional state into consideration.

[0573] A "user" is an entity that uses the system to input an implementation plan for advertising measures and receives forecast results of sales trends and proposals for optimizing advertising measures.

[0574] A "terminal" is a device that allows a user to input an implementation plan for an advertising campaign and transmit that information to a server.

[0575] The "server" is a computer system that receives information sent by users, performs preprocessing, and uses generative AI models and emotion engines to predict sales trends and propose optimized advertising strategies.

[0576] A "generative AI model" is an artificial intelligence model that predicts sales trends based on advertising campaign information entered by the user.

[0577] A "prompt" is an instruction entered into the generative AI model to predict sales trends.

[0578] The "emotion engine" is an artificial intelligence engine that analyzes the user's emotional state and uses that information to suggest ways to optimize advertising strategies.

[0579] The "advertising campaign implementation plan" is specific planning information entered by the user, such as the advertisement distribution date and time, region, advertising cost, and type of advertisement.

[0580] "Sales trends" is data indicating predicted fluctuations in sales after the implementation of advertising measures.

[0581] "Preprocessing" is data processing to convert information received from users into a format suitable for the generative AI model.

[0582] The "prediction result" is the predicted data of sales trends that the generative AI model outputs based on the prompt sentence.

[0583] "Optimization suggestions" are specific measures proposed by the emotion engine based on the user's emotional state to maximize the effectiveness of advertising campaigns.

[0584] MODE FOR CARRYING OUT THE INVENTION

[0585] This invention is a system that allows a user to input an implementation plan for an advertising campaign, predicts sales trends based on that information, and also proposes optimization of the advertising campaign taking into account the emotional state of the user. A specific embodiment of this system is described below.

[0586] System configuration

[0587] The system consists of a terminal where users can input their advertising campaign implementation plans, a server that receives and processes the input information, a generative AI model that predicts sales trends, and an emotion engine that analyzes the user's emotional state.

[0588] Hardware and software used

[0589] Terminal: A device (e.g., PC, smartphone) through which a user inputs an advertising implementation plan.

[0590] Server: The computer system that receives information, performs preprocessing, and runs the generative AI model and emotion engine.

[0591] Generative AI models: Built using machine learning libraries such as TensorFlow and PyTorch.

[0592] Emotion Engine: Built using natural language processing models such as OpenAI's GPT-3 (registered trademark) and BERT.

[0593] Data processing and calculation

[0594] 1. The user enters an advertising implementation plan: The user accesses the system's web interface and enters in a text box, "Next Monday, in the Tokyo area, we will distribute Ad A with an advertising budget of 50,000 yen."

[0595] 2. The terminal sends the entered information to the server: The terminal sends the information entered by the user to the server as an HTTP request. For example, it sends the information using a POST request.

[0596] 3. The server preprocesses the received information: The server parses the received information in JSON format, converts the date and time information into "YYYY-MM-DD" format, and converts the region information into a geographic code (e.g., Tokyo = 13).

[0597] 4. The server inputs a prompt into the generative AI model to predict sales trends: Based on the preprocessed information, the server generates a prompt such as "Please predict sales trends if Ad A is distributed in the Tokyo area next Monday with an advertising budget of 50,000 yen," and inputs it into the generative AI model. The generative AI model is built using TensorFlow.

[0598] 5. The server presents the prediction results to the user: The server generates the sales trend prediction results obtained from the generative AI model in HTML format and displays them in the user's browser. For example, it visually displays the predicted sales trend using a line graph.

[0599] 6. The server collects the user's emotional state: The server displays a questionnaire form asking the user about their emotional state. The user selects their emotional state from the options and clicks the submit button.

[0600] 7. The server uses an emotion engine to suggest optimization of advertising measures: The server uses an emotion engine (e.g., OpenAI's GPT-3) based on the collected emotion data to suggest optimization of advertising measures. For example, it generates a result such as "If a user sees a specific advertisement A when they are happy, it determines that advertisement A is likely to be well-received by the user, and suggests a similar advertising measure," and presents it to the user.

[0601] Examples of specific examples and prompts

[0602] For example, a user might enter the following advertising plan:

[0603] "Next Monday, we will distribute Ad A in the Tokyo area with an advertising budget of 50,000 yen."

[0604] In this case, the server inputs the following prompt sentence into the generative AI model:

[0605] "Please predict the sales trend if Ad A is distributed in the Tokyo area next Monday with an advertising budget of 50,000 yen."

[0606] And the following prompt for the emotion engine:

[0607] "If a user sees a specific ad A when they are happy, determine that ad A is likely to be well-received by the user and suggest a similar advertising strategy."

[0608] In this way, the server provides the user with a sales trend forecast result and an optimized advertising measure. The flow of the identification process in the third embodiment will be described with reference to FIG.

[0609] Step 1:

[0610] The user inputs an implementation plan for the advertising campaign.

[0611] Input: The user enters details of the advertising campaign (e.g., "Next Monday, distribute Ad A in the Tokyo area with an advertising budget of 50,000 yen") into the system's web interface.

[0612] Output: Implementation plan data for the entered advertising campaign.

[0613] Specific operation: The user accesses the system's web interface, enters details of the advertising campaign in the text box, and clicks the submit button.

[0614] Step 2:

[0615] The terminal transmits the input information to the server.

[0616] Input: Advertising implementation plan data entered by the user.

[0617] Output: Advertising implementation plan data sent to the server.

[0618] Specific operation: The device sends the information entered by the user to the server as an HTTP POST request, which includes details of the advertising campaign.

[0619] Step 3:

[0620] The server preprocesses the information it receives.

[0621] Input: Advertising implementation plan data sent from the device.

[0622] Output: Preprocessed advertising campaign implementation plan data.

[0623] Specific operation: The server parses the received information in JSON format, converts the date and time information to "YYYY-MM-DD" format, converts the region information to a geographic code (e.g., Tokyo = 13), and checks for missing data and completes it if necessary.

[0624] Step 4:

[0625] The server inputs prompt statements into the generative AI model to predict sales trends.

[0626] Input: Preprocessed advertising campaign implementation plan data.

[0627] Output: Sales trend prediction results from the generative AI model.

[0628] Specific operation: Based on the preprocessed information, the server generates a prompt statement such as "Please predict the sales trend if Ad A is distributed in the Tokyo area next Monday with an advertising budget of 50,000 yen," and inputs it into the generative AI model. The generative AI model is built using TensorFlow.

[0629] Step 5:

[0630] The server presents the prediction results to the user.

[0631] Input: Sales trend prediction results obtained from the generative AI model.

[0632] Output: Sales forecast results presented to the user.

[0633] Specific operation: The server generates the sales trend prediction results obtained from the generative AI model in HTML format and displays them in the user's browser. For example, it visually displays the predicted sales trend using a line graph.

[0634] Step 6:

[0635] The server collects the user's emotional state.

[0636] Input: Data about the user's emotional state.

[0637] Output: Collected user emotional state data.

[0638] Specific operation: The server displays a questionnaire form asking the user about their emotional state. The user selects their emotional state from the options and clicks the submit button.

[0639] Step 7:

[0640] The server uses an emotion engine to propose optimization of advertising strategies.

[0641] Input: Collected user emotional state data.

[0642] Output: Optimization proposals for advertising measures using an emotion engine.

[0643] Specific operation: The server uses an emotion engine (e.g., OpenAI's GPT-3) based on the collected emotion data to propose optimization of advertising measures. For example, it generates a result such as "If a user sees a specific ad A when they are happy, it determines that ad A is likely to be well-received by the user, and suggests a similar advertising measure," and presents it to the user.

[0644] (Application example 3)

[0645] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0646] Conventional post-sales analysis systems for internet advertising have issues with predicting sales trends based on advertising implementation plans and not adequately optimizing advertising measures by taking into account the emotional state of users. This makes it difficult to maximize advertising effectiveness and may not meet advertisers' expectations.

[0647] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 3 is realized by the following means. In this invention, the server includes means for learning sales trends after the implementation of a campaign and outputting an estimated result of the sales trend, means for presenting the estimated result to the user, and means for analyzing the emotional state of the user and optimizing the advertising campaign. This makes it possible to predict sales trends based on an implementation plan of the advertising campaign and to optimize the advertising campaign taking into account the emotional state of the user.

[0648] "Internet advertising" refers to advertising delivered via the Internet.

[0649] "Sales analysis" refers to analyzing sales data obtained after implementing advertising measures and evaluating their effectiveness.

[0650] "AI" stands for artificial intelligence, a technology that mimics human intelligence through machine learning and data analysis.

[0651] "Sales trends" refers to data showing fluctuations in sales over a certain period of time.

[0652] "Predicted results" refers to data that shows future sales prospects predicted by AI.

[0653] "User" refers to advertisers and marketers who use the system.

[0654] "Emotional state" refers to the user's psychological state or feelings.

[0655] An "advertising strategy" is an advertising initiative designed to achieve a specific objective.

[0656] "Optimization" means adjusting conditions or parameters to obtain the best results in order to achieve a specific objective.

[0657] "Learning" is the process by which AI finds patterns and rules based on past data and uses them to make future predictions and decisions.

[0658] "Presentation" means that the system displays information to the user.

[0659] A system for carrying out this invention includes an AI specialized in analyzing sales after the implementation of internet advertising. The system includes means for learning sales trends after the implementation of a campaign and outputting predicted results of the sales trends, means for presenting the predicted results to a user, and means for analyzing the emotional state of the user and optimizing the advertising campaign.

[0660] 1. System Program

[0661] The system executes the following program:

[0662] Hardware: Smartphone

[0663] Software: Python, scikit-learn, sentiment analysis library (e.g., TextBlob), ad performance prediction library

[0664] 2. Program processing explanation

[0665] The server uses a sales forecasting model to predict sales trends based on the advertising campaign data entered by the user. Specifically, if a user enters a plan to "distribute Ad A in the Tokyo area next Monday with an advertising budget of 50,000 yen," the AI ​​will use this information to predict sales trends and present the results to the user.

[0666] Furthermore, the server uses an emotion engine to analyze the user's emotional state. For example, if the user is in a "happy" emotional state, the server optimizes advertising measures based on that emotional state. Specifically, if the user sees a specific advertisement A while feeling happy, the server determines that advertisement A is likely to be well-received by the user and suggests similar advertising measures.

[0667] 3. Examples and prompts

[0668] As a concrete example, consider the case where an advertiser inputs a plan to "distribute Ad A in the Tokyo area next Monday with an advertising budget of 50,000 yen." In this case, the system generates the following prompt sentence:

[0669] Example prompt sentence:

[0670] Enter your advertising strategy:

[0671] Region: Tokyo

[0672] Advertising cost: 50,000 yen

[0673] Ad Type: Ad A

[0674] Date: 2023-10-09

[0675] User's emotional state: joy

[0676] When this prompt sentence is input into the generative AI model, it outputs sales forecasts and optimized advertising measures, enabling advertisers to implement effective advertising measures.

[0677] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[0678] Step 1:

[0679] The user inputs the advertising plan, specifically the information such as region, advertising cost, advertising type, date, user's emotional state, etc. The input data is sent to the server.

[0680] input:

[0681] Region: Tokyo

[0682] Advertising cost: 50,000 yen

[0683] Ad Type: Ad A

[0684] Date: 2023-10-09

[0685] User's emotional state: joy

[0686] output:

[0687] Advertising campaign plan data sent to the server.

[0688] Step 2:

[0689] The server uses a sales forecasting model to predict sales trends based on the received advertising campaign plan data. Specifically, it uses past advertising execution results data and applies a machine learning algorithm (e.g., Linear Regression) to predict future sales trends.

[0690] input:

[0691] Advertising campaign planning data, past advertising results data.

[0692] output:

[0693] Projected sales trend data.

[0694] Step 3:

[0695] The server presents the predicted sales trend data to the user. Specifically, it generates sales trend graphs and numerical data and displays them on the user's terminal.

[0696] input:

[0697] Projected sales trend data.

[0698] output:

[0699] Sales trend graphs and numerical data displayed on the user's device.

[0700] Step 4:

[0701] The server analyzes the user's emotional state using an emotion engine. Specifically, it calculates an emotion score using an emotion analysis library (e.g., TextBlob) based on the emotional state data entered by the user.

[0702] input:

[0703] User emotional state data.

[0704] output:

[0705] The calculated sentiment score.

[0706] Step 5:

[0707] The server optimizes advertising measures based on the emotional score. Specifically, if the emotional score is high, it will suggest similar advertising measures, and if the emotional score is low, it will suggest changing the advertising type or region.

[0708] input:

[0709] Calculated sentiment scores and advertising strategy planning data.

[0710] output:

[0711] Proposal of optimized advertising measures.

[0712] Step 6:

[0713] The server presents the optimized advertising proposal to the user. Specifically, the proposal is generated in text format and displayed on the user's device.

[0714] input:

[0715] Proposal of optimized advertising measures.

[0716] output:

[0717] Proposing optimized advertising measures to be displayed on the user's device.

[0718] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0719] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0720] Another example of generative AI is Gemini (registered trademark) (Internet search engine). <url: https: gemini.google.com ?hl="ja">) are listed.

[0721] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0722] [Second embodiment]

[0723] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0724] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0725] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0726] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0727] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0728] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0729] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0730] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0731] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0732] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0733] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0734] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[0735] "Example 1"

[0736] An example of the present invention is a system including AI specialized in analyzing sales after the implementation of internet advertising. This system collects sales data after the implementation of advertising measures, and the AI ​​learns from this data. The AI ​​predicts future sales trends based on past advertising results and sales data. The prediction results are output and presented to the user in a format that is easy for the user to understand.

[0737] "Example 2"

[0738] Specifically, AI uses deep learning models to learn and predict sales trends. Past advertising results and sales data include information such as the type of ad, ad delivery time, delivery area, and advertising costs. Based on this information, AI predicts sales trends after the implementation of advertising measures.

[0739] "Example 3"

[0740] More specifically, when a user inputs an implementation plan for an advertising campaign, the AI ​​predicts sales trends based on this plan. For example, if a user inputs a plan such as "Next Monday, in the Tokyo area, we will distribute Ad A with an advertising budget of 50,000 yen," the AI ​​will predict sales trends based on this information and present the results to the user.

[0741] The processing flow of each embodiment will be described below.

[0742] "Example 1"

[0743] Step 1: After the advertising campaign is implemented, the system collects sales data.

[0744] Step 2: The collected sales data is learned by AI.

[0745] Step 3: The AI ​​predicts future sales trends based on past advertising results and sales data.

[0746] Step 4: The prediction results are output in a user-friendly format and presented to the user.

[0747] "Example 2"

[0748] Step 1: The AI ​​uses a deep learning model to learn sales trends.

[0749] Step 2: Past advertising results and sales data includes information such as the type of ad, ad delivery time, delivery area, and advertising costs.

[0750] Step 3: Based on this information, the AI ​​predicts sales trends after the advertising campaign is implemented.

[0751] "Example 3"

[0752] Step 1: The user enters an implementation plan for their advertising campaign.

[0753] Step 2: The AI ​​predicts sales trends based on this plan.

[0754] Step 3: For example, if a user inputs a plan such as "Next Monday, in the Tokyo area, we will distribute Ad A with an advertising budget of 50,000 yen," the AI ​​will use this information to predict sales trends.

[0755] Step 4: The prediction results are presented to the user.

[0756] Example 1

[0757] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0758] Quantitatively evaluating the effectiveness of internet advertising and accurately predicting future sales trends is important for maximizing the effectiveness of advertising measures. However, with conventional methods, it was difficult to efficiently collect and analyze sales data after the implementation of advertising measures and predict future sales. Another challenge was presenting the prediction results in a format that was easy for users to understand.

[0759] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0760] In this invention, the server includes means for collecting sales data after the implementation of internet advertising, means for preprocessing the collected sales data, and means for training an artificial intelligence model using the preprocessed data, thereby making it possible to quantitatively evaluate the effectiveness of advertising measures and accurately predict future sales trends.

[0761] "Internet advertising" refers to advertising displayed on the Internet and delivered to users through websites, social media, search engines, etc.

[0762] "Sales data" refers to data relating to the sale of goods and services, and includes information such as the sales amount, sales quantity, and sales date and time.

[0763] "Artificial intelligence" refers to technologies that allow computer systems to learn, reason, and self-correct by imitating human intelligence, and includes machine learning and deep learning.

[0764] "Data collection means" refers to the methods and technologies for obtaining sales data after the implementation of internet advertising, including data acquisition through APIs.

[0765] "Data preprocessing means" refers to methods and techniques for preparing collected sales data in an analyzable format, including filling in missing values, removing outliers, and normalizing data.

[0766] An "artificial intelligence model" is a mathematical model that learns from data and makes predictions and classifications, and includes neural networks and regression models.

[0767] "Learning tools" refers to methods and techniques for training artificial intelligence models using pre-processed data, including machine learning algorithms and deep learning frameworks.

[0768] "Prediction means" refers to methods and technologies for predicting future sales trends using a trained artificial intelligence model, and includes the model's inference function.

[0769] "Presentation means" refers to the method or technique for visually displaying the prediction results to the user, including graphical and tabular presentations.

[0770] The present invention is a system including artificial intelligence specialized for analyzing sales after the execution of internet advertising. Specific embodiments of this system will be described below.

[0771] The server collects sales data after the online advertisements are executed. Specifically, the sales data is obtained through an API. For example, the sales data can be obtained using a web service API. The server uses the Python Requests library to send a request to the API endpoint and obtain the sales data.

[0772] Next, the server preprocesses the collected sales data, specifically by imputing missing values, removing outliers, normalizing the data, etc. The server uses the Pandas library to create a data frame, impute missing values, remove outliers, and normalize the data.

[0773] The server then uses the preprocessed data to train an artificial intelligence model. Specifically, it uses TensorFlow and PyTorch to build a model based on past advertising results and sales data. The server then uses TensorFlow to build a neural network model and trains it on the data.

[0774] After the learning is complete, the server uses the learned artificial intelligence model to predict future sales trends. Specifically, it predicts sales for the next three months. The server uses the trained model to predict sales for the next three months.

[0775] Finally, the terminal presents the prediction results received from the server to the user in the form of graphs and tables. The terminal uses the Matplotlib library to graph the prediction results and display them to the user.

[0776] As a concrete example, the following prompt sentence can be input to a generative AI model:

[0777] Example prompt sentence:

[0778] Collect sales data after implementing advertising campaign A and train the AI ​​model. Based on the advertising campaign and sales data from the past year, predict sales trends for the next three months.

[0779] In this way, users can quantitatively evaluate the effectiveness of their internet advertising and predict future sales. This system enables data-driven decision-making to maximize the effectiveness of advertising initiatives.

[0780] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0781] Step 1:

[0782] The server collects sales data after the implementation of online advertising. As input, it uses an API endpoint for obtaining sales data after the implementation of advertising campaigns. Specifically, the server uses the Python Requests library to send a request to the API and obtains the sales data in JSON format. The output is the obtained sales data.

[0783] Step 2:

[0784] The server preprocesses the collected sales data. It uses the sales data obtained in step 1 as input. Specifically, the server creates a data frame using the Pandas library, imputes missing values, removes outliers, and normalizes the data. The output is the preprocessed sales data.

[0785] Step 3:

[0786] The server uses the preprocessed data to train an artificial intelligence model. The sales data preprocessed in step 2 is used as input. Specifically, the server uses TensorFlow to build a neural network model and trains it on the data. The output is a trained artificial intelligence model.

[0787] Step 4:

[0788] The server uses the trained artificial intelligence model to predict future sales trends. As input, it uses the trained artificial intelligence model and data for the prediction period. Specifically, the server uses the trained model to predict sales for the next three months. The output is the predicted sales data.

[0789] Step 5:

[0790] The terminal presents the forecast results received from the server to the user. As input, it uses the sales data predicted in step 4. Specifically, the terminal uses the Matplotlib library to graph the forecast results and display them to the user. The output is the forecast results presented to the user.

[0791] (Application example 1)

[0792] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0793] Accurately analyzing the effectiveness of internet advertising and predicting future sales trends are important challenges for advertisers. However, conventional methods have made it difficult to analyze advertising campaign data in real time and forecast sales. There has also been a lack of ways to present the forecast results in a format that is easy for users to understand. This has made it difficult to make appropriate decisions to maximize the effectiveness of advertising measures.

[0794] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0795] In this invention, the server is a system including an AI specialized in analyzing sales after the implementation of internet advertising. The AI ​​includes means for learning sales trends after the implementation of a campaign and outputting predicted results of sales trends, means for presenting the predicted results to users, means for inputting advertising campaign data and making sales forecasts in real time, and means for displaying the predicted results in graphs and charts. This makes it possible to analyze advertising campaign data in real time and make sales forecasts. In addition, visually displaying the predicted results makes them easy for users to understand and supports appropriate decision-making.

[0796] "Internet advertising" refers to advertising displayed on the Internet, a means of providing information to users through websites and applications.

[0797] "Sales analysis" is the process of collecting and analyzing sales data over a specific period of time to identify sales trends and factors.

[0798] "AI" is an abbreviation for artificial intelligence, a technology that allows computers to imitate human intelligence to learn and reason.

[0799] "Sales trends after the implementation of the campaign" refers to the fluctuations in sales over a certain period of time after the implementation of the advertising campaign.

[0800] The "predicted results" are future sales forecasts based on the data learned by the AI.

[0801] "User" refers to any individual or legal entity that uses the System.

[0802] An "advertising campaign" is a planned advertising activity with a specific purpose, and may include multiple advertising initiatives.

[0803] "Real-time" refers to data being processed and analyzed as it is generated.

[0804] "Graphs and charts" are diagrams that visually represent numerical data, and include bar graphs, line graphs, pie charts, etc.

[0805] A "generative AI model" is an artificial intelligence model trained to perform a specific task, such as making predictions or classifications based on data.

[0806] A "prompt sentence" is a sentence that gives instructions or asks questions to the user, and serves to guide the operation of the system.

[0807] As an embodiment of the present invention, an advertising effectiveness analysis assistant system will be specifically described.

[0808] System configuration

[0809] This system includes AI specialized in analyzing sales after the execution of internet advertising. The system consists of the following main components:

[0810] 1. Data collection module: collects advertising campaign data and sends it to the server.

[0811] 2. Data Preprocessing Module: Cleans the collected data and converts it into a suitable format for analysis.

[0812] 3. AI model: Learns from past advertising results and sales data and predicts future sales trends.

[0813] 4. Prediction result display module: Visually displays the prediction results in graphs and charts.

[0814] Hardware and software used

[0815] Hardware: Smartphones, servers

[0816] Software: Python, pandas, scikit-learn, matplotlib

[0817] Data processing and calculation

[0818] The server processes the data in the following steps:

[0819] 1. Data Loading: The data collection module sends the advertising campaign data to the server as a CSV file, which the server loads.

[0820] 2. Data Preprocessing: The data preprocessing module removes missing values ​​and cleans the data.

[0821] 3. Model training: The AI ​​model trains a linear regression model targeting sales using ad spend, clicks, and impressions as features.

[0822] 4. Sales forecasting: The AI ​​model predicts sales based on new data.

[0823] 5. Displaying the results: The prediction result display module displays the prediction results as graphs and charts.

[0824] Specific examples

[0825] Users prepare their ad campaign data as a CSV file and upload it to the app. Then, they input the data for their new ad campaign (advertising costs, clicks, impressions). The server uses this data to generate sales forecasts and displays the results in graphs.

[0826] Prompt Sentence Examples

[0827] "Upload your ad campaign data and enter new campaign data. The AI ​​will generate sales forecasts and display the results in graphs."

[0828] In this way, the advertising effectiveness analysis assistant system can analyze advertising campaign data in real time and make sales forecasts. In addition, by visually displaying the forecast results, users can easily understand them and help them make appropriate decisions.

[0829] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0830] Step 1:

[0831] The user prepares advertising campaign data as a CSV file and uploads it to the app from their device. The input is the advertising campaign data, and the output is the CSV file sent to the server. Specifically, the user clicks the "Upload Data" button in the app and selects the CSV file from the file selection dialog.

[0832] Step 2:

[0833] The server reads the uploaded CSV file. The input is a CSV file, and the output is the advertising campaign data in a data frame. Specifically, the server converts the CSV file into a data frame using the Python pandas library.

[0834] Step 3:

[0835] The server uses a data preprocessing module to clean the data. The input is the advertising campaign data in a data frame format, and the output is the clean data with missing values ​​removed. Specifically, the server removes missing values ​​and converts the data into the required data type.

[0836] Step 4:

[0837] The server uses the AI ​​model to learn from past advertising results and sales data. The input is clean data, and the output is a trained AI model. Specifically, the server trains a linear regression model using the scikit-learn library.

[0838] Step 5:

[0839] The user inputs data for a new advertising campaign (advertising cost, clicks, impressions) into the app from their device. The input is the new advertising campaign data, and the output is the new data sent to the server. Specifically, the user enters the data into the app's input form and clicks the "Start prediction" button.

[0840] Step 6:

[0841] The server uses the trained AI model to make sales predictions based on new data. The input is new advertising campaign data, and the output is the sales prediction results. Specifically, the server inputs new data into the AI ​​model and obtains the prediction results.

[0842] Step 7:

[0843] The server uses the forecast result display module to visually display the forecast results in graphs and charts. The input is the sales forecast results, and the output is the graphs and charts presented to the user. Specifically, the server uses the matplotlib library to plot the forecast results on a graph and display it on the user's terminal.

[0844] The above are the specific processing steps of the advertising effectiveness analysis assistant system.

[0845] Example 2

[0846] Next, a description will be given of Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0847] Conventional internet advertising effectiveness analysis systems have difficulty accurately predicting sales trends after implementing advertising measures. In addition, there is a lack of effective ways to utilize past advertising results, and highly accurate predictions are required when planning advertising measures.

[0848] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0849] In this invention, the server includes means for collecting past advertising implementation results and sales data, means for preprocessing the collected data, means for training a deep learning model using the preprocessed data, means for a user to input information on new advertising measures, means for predicting sales trends based on the input information on the advertising measures, and means for presenting the prediction results to the user. This makes it possible to predict sales trends after the implementation of advertising measures with high accuracy and provide useful information to the user.

[0850] "Past advertising results" refers to detailed data when advertising was carried out, and includes information such as the type of advertisement, distribution time, distribution area, and advertising costs.

[0851] "Sales data" refers to data that indicates the sales performance of a product during a specific period, and includes information such as sales amount and sales quantity.

[0852] "Preprocessing" refers to the process of converting collected data into a format suitable for analysis and model training, and includes imputing missing values, normalizing data, encoding categorical data, etc.

[0853] A "deep learning model" is an algorithm that uses a multi-layer neural network to learn the characteristics of data and perform predictions and classifications.

[0854] "User" refers to an individual or corporation that uses the system to predict the effectiveness of advertising measures.

[0855] "Information about advertising measures" is detailed data about a new advertisement planned by the user, and includes information such as the type of advertisement, distribution time, distribution area, and advertising cost.

[0856] "Sales trend forecasting" refers to predicting fluctuations in sales after the implementation of advertising campaigns, and is done using a deep learning model.

[0857] The "prediction result" is a predicted value of sales trends calculated by a deep learning model and is presented to the user.

[0858] This invention is a system that predicts sales trends after implementing advertising measures based on past advertising results and sales data. The system is composed of three main elements: a server, a terminal, and a user.

[0859] server

[0860] The server runs on the following hardware and software:

[0861] Hardware: A server machine equipped with a high-performance CPU, GPU, memory, and storage

[0862] Software: Deep learning frameworks such as TensorFlow and PyTorch, database management systems (e.g., MySQL, PostgreSQL)

[0863] The server first collects past advertising results and sales data, which involves retrieving data from the company's database or external APIs, and then stores the collected data in a database.

[0864] The server then preprocesses the collected data, specifically by imputing missing values, normalizing the data, encoding categorical data, etc. The preprocessed data is then converted into a format suitable for input into a deep learning model.

[0865] The server uses the preprocessed data to train a deep learning model. Training is performed using TensorFlow and PyTorch, and the model parameters are optimized. Once trained, the model is used to predict sales trends after the implementation of advertising campaigns.

[0866] Terminal

[0867] The terminal provides an interface for users to input information for new advertising campaigns. The terminal operates using the following hardware and software:

[0868] Hardware: PCs, tablets, smartphones, etc.

[0869] Software: Web browser, dedicated application

[0870] The user inputs information about the new advertising campaign into an input form on the device, including the type of advertisement, distribution time, distribution area, and advertising cost.

[0871] User

[0872] Users are individuals or corporations who use the system to predict the effectiveness of advertising campaigns. Users input information about new advertising campaigns through their terminals and receive prediction results.

[0873] Specific examples

[0874] For example, if a user inputs "A new TV ad will be broadcast in Tokyo from 8:00 PM to 10:00 PM, with an advertising cost of 1 million yen," the server inputs this information into the deep learning model and calculates the predicted sales trend. The server sends the calculation results to the terminal and displays to the user, "Predicted sales trend: 10% increase after one week, 5% increase after one month."

[0875] Prompt Sentence Examples

[0876] "Please predict the sales trend if a new TV ad is broadcast in Tokyo from 8:00 PM to 10:00 PM and the advertising cost is 1 million yen."

[0877] In this way, the server uses a deep learning model to predict the effectiveness of advertising campaigns and provide them to the user.

[0878] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0879] Step 1: Data collection

[0880] The server collects past advertising results and sales data from the company's database or external API. As input, it obtains information such as the type of ad, delivery time, delivery area, and advertising cost. As output, it stores this data in the database. Specifically, the server sends an API request and stores the obtained data in the database.

[0881] Step 2: Data Preprocessing

[0882] The server preprocesses the collected data. As input, it uses the advertising performance results and sales data stored in the database. As output, it generates the preprocessed data. Specifically, the server performs missing value imputation, data normalization, and categorical data encoding. For example, if there are missing values, it imputes them with the mean value, and standardizes numerical data.

[0883] Step 3: Model training

[0884] The server trains a deep learning model using the preprocessed data. It uses the preprocessed data as input and generates a trained deep learning model as output. Specifically, the server uses TensorFlow or PyTorch to build a neural network and optimizes the model parameters using the training data. For training, it uses the Adam optimizer or cross-entropy loss function.

[0885] Step 4: Enter a new advertising strategy

[0886] The user inputs information about a new advertising campaign into an input form on the device. As input, the user provides information such as the type of advertisement, distribution time, distribution area, and advertising cost. As output, the entered information about the advertising campaign is sent to the server. Specifically, the user inputs the information using a web browser or a dedicated application and clicks the send button.

[0887] Step 5: Prediction execution

[0888] The server inputs the information of new advertising campaigns entered by the user into the deep learning model and performs predictions. As input, it uses the information of advertising campaigns provided by the user. As output, it generates predicted sales trends. Specifically, the server inputs the information of advertising campaigns into the deep learning model, and the model calculates sales trends.

[0889] Step 6: View the results

[0890] The server sends the prediction results to the terminal and displays them to the user. The predicted sales trend is used as input. The prediction results are displayed to the user as output. Specifically, the server sends the prediction results to the terminal, and the terminal displays the results on the screen. For example, it displays results such as "Predicted sales trend: 10% increase after one week, 5% increase after one month."

[0891] (Application example 2)

[0892] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0893] Conventional internet advertising effectiveness analysis systems have difficulty accurately predicting sales trends after implementing advertising measures, and do not provide sufficient support for advertisers to develop optimal advertising strategies. Furthermore, there is no way for advertisers to easily predict the effectiveness of their advertising campaigns using smart devices. This has made it difficult for advertisers to optimally allocate advertising costs and implement effective advertising measures.

[0894] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0895] In this invention, the server includes means including an AI specialized in analyzing sales after the implementation of internet advertising, means for the AI ​​to learn sales trends after the implementation of a campaign and output predicted sales trends, means for presenting the predicted results to a user, and means for a user to predict the effectiveness of an advertising campaign using a smart device. This enables advertisers to learn sales trends based on past advertising results and predict sales trends based on information such as the type of advertisement, distribution time, distribution area, and advertising costs. In addition, advertisers can input data for a new advertising campaign using a smart device and predict sales, thereby receiving support for formulating an optimal advertising strategy.

[0896] "Internet advertising" refers to advertising distributed over the Internet, displayed to users through websites, social media, search engines, etc.

[0897] "Sales analysis" is the process of analyzing sales data obtained after implementing advertising measures and evaluating their effectiveness.

[0898] "AI" stands for artificial intelligence, which is a computer system designed to perform specific tasks automatically.

[0899] "Sales trends after the implementation of the campaign" refers to the fluctuations in sales over a certain period of time after the implementation of the advertising campaign.

[0900] The "predicted results" are future sales prospects predicted by AI based on past data.

[0901] "Users" refer to advertisers, marketers, and other people who use this system to predict advertising effectiveness.

[0902] A "smart device" is a portable electronic device with internet connectivity, such as a smartphone or tablet.

[0903] An "advertising campaign" is a series of advertising activities designed to achieve a specific objective.

[0904] "Ad type" refers to the format or media type of the ad, such as banner ad, video ad, or text ad.

[0905] "Serving time" refers to the time period during which an advertisement is displayed to a user.

[0906] "Distribution Area" means the geographic area in which an advertisement is displayed.

[0907] "Advertising costs" are expenses necessary to carry out an advertising campaign.

[0908] "Learning sales trends" is the process of building a model to predict future sales fluctuations based on past sales data.

[0909] "Predicting sales trends" means estimating future sales fluctuations based on past data.

[0910] An "advertising implementation plan" is a plan that includes the specific implementation details and schedule of an advertising campaign.

[0911] A system for carrying out the present invention includes an AI specialized for analyzing sales after the execution of internet advertising. A specific embodiment of this system will be described below.

[0912] System configuration

[0913] The system consists of the following main components:

[0914] 1. Server: Responsible for processing data and running AI models.

[0915] 2. Smart devices: Used by users to predict the effectiveness of advertising campaigns.

[0916] 3. Database: Stores past advertising results and sales data.

[0917] Hardware and software used

[0918] Hardware: Servers, smartphones, tablets

[0919] Software: Python, Pandas, Scikit-learn, TensorFlow / Keras

[0920] Data processing and calculation

[0921] The server processes the data in the following steps to predict sales trends.

[0922] 1. Data loading: The server loads past advertising results and sales data from the database.

[0923] 2. Data preprocessing: Using Pandas, categorical data such as ad type, delivery time, delivery area, and ad cost are converted into dummy variables, and the data is standardized using Scikit-learn.

[0924] 3. Building and training the model: Build a deep learning model using TensorFlow / Keras and train it with the training data.

[0925] 4. Sales forecasting: Input data for a new advertising campaign and use the AI ​​model to predict sales.

[0926] Specific examples

[0927] Forecast sales if an advertiser distributes a new banner ad in Tokyo at 12:00 and spends 50,000 yen on the ad.

[0928] Prompt Sentence Examples

[0929] What is the sales forecast for a new banner ad distributed in Tokyo at 12:00 with an advertising cost of 50,000 yen?

[0930] Processing flow

[0931] 1. User input: The user uses a smart device to input data for a new advertising campaign (ad type, delivery time, delivery area, advertising cost).

[0932] 2. Data transmission: Data is sent from the smart device to the server.

[0933] 3. Sales forecast: The server runs an AI model based on the received data to predict sales trends.

[0934] 4. Presenting the results: The server sends the prediction results to the smart device and presents them to the user.

[0935] In this way, advertisers can predict future sales based on past data and receive support in formulating optimal advertising strategies.

[0936] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0937] Step 1:

[0938] User Input

[0939] The user uses a smart device to enter data for a new advertising campaign (ad type, distribution time, distribution area, advertising cost), which is then formatted by the smart device application and prepared for transmission to the server.

[0940] Input: Ad type, delivery time, delivery area, ad cost

[0941] Output: Formatted ad campaign data

[0942] Step 2:

[0943] Sending data

[0944] The smart device sends the advertising campaign data entered by the user to a server, where it is securely transmitted over the internet.

[0945] Input: Formatted ad campaign data

[0946] Output: Ad campaign data sent to the server

[0947] Step 3:

[0948] Loading data

[0949] The server reads past advertising results and sales data from the database, including information on the type of ad, distribution time, distribution area, advertising cost, and sales.

[0950] Input: Past advertising results and sales data from the database

[0951] Output: Loaded past advertising results and sales data

[0952] Step 4:

[0953] Data Preprocessing

[0954] The server uses Pandas to convert categorical data such as ad type, delivery time, delivery area, and ad cost into dummy variables, and then uses Scikit-learn to standardize the data, preparing it in a format suitable for the AI ​​model.

[0955] Input: Loaded past advertising results and sales data

[0956] Output: Preprocessed data (dummy variables and standardized data)

[0957] Step 5:

[0958] Building and training the model

[0959] The server builds a deep learning model using TensorFlow / Keras and trains the model using the preprocessed data. The trained model has the ability to predict future sales based on past data.

[0960] Input: Preprocessed data

[0961] Output: A trained deep learning model

[0962] Step 6:

[0963] Sales forecast

[0964] The server preprocesses new ad campaign data submitted by users and uses the trained deep learning model to predict sales, outputting the results as a numerical value indicating the likelihood of future sales.

[0965] Input: New advertising campaign data, trained deep learning model

[0966] Output: Sales forecast results

[0967] Step 7:

[0968] Presentation of results

[0969] The server sends the prediction results to the smart device and displays them to the user, who can then check the results on the screen and use them to plan their advertising strategies.

[0970] Input: Sales Forecast Results

[0971] Output: Prediction results displayed on a smart device

[0972] In this way, advertisers can predict future sales based on past data and receive support in formulating optimal advertising strategies.

[0973] Example 3

[0974] Next, a description will be given of Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0975] With conventional sales analysis systems after the implementation of internet advertising, it was difficult to predict sales trends based on the implementation plan of advertising measures. In addition, there was a lack of means to effectively utilize past advertising results to predict sales trends, making it difficult to evaluate the effectiveness of advertising measures in advance. This made it difficult to optimally allocate advertising expenses and plan effective advertising measures.

[0976] The identification process by the identification processing unit 290 of the data processing device 12 in Example 3 is realized by the following means. In this invention, the server includes a means for a user to input an implementation plan for an advertising measure, a means for a terminal to transmit the input plan to the server, a means for the server to input plan information to the generative AI model, a means for the generative AI model to predict sales trends, a means for the server to receive and transmit the prediction results to the terminal, and a means for the terminal to present the prediction results to the user. This enables the user to predict sales trends based on the implementation plan for the advertising measure and effectively utilize past advertising implementation results.

[0977] A "user" is an entity that uses the system to input an implementation plan for advertising measures and receives forecast results for sales trends.

[0978] A "terminal" is a device through which a user inputs an implementation plan for an advertising campaign and communicates with a server to receive predicted results.

[0979] The "server" is a computer system that receives planning information sent by the user, inputs it into the generative AI model, generates prediction results, and sends them to the terminal.

[0980] A "generative AI model" is an artificial intelligence model that predicts sales trends based on an advertising implementation plan, and has the ability to learn from past advertising implementation results.

[0981] The "advertising measure implementation plan" is planning information including the details and conditions of the advertisement implementation input by the user.

[0982] "Sales trends" is data indicating predicted fluctuations in sales after the implementation of advertising measures.

[0983] "Prediction results" are sales trend forecast data calculated by the generative AI model based on the implementation plan of the advertising campaign.

[0984] This invention is a system that allows a user to input an implementation plan for an advertising campaign and predicts sales trends based on the plan. The system includes a user, a terminal, a server, and a generative AI model.

[0985] The user inputs an implementation plan for advertising measures through the device interface. For example, the user might input a plan such as "Next Monday, in the Tokyo area, we will distribute Ad A with an advertising budget of 50,000 yen." The device receives this input and sends the plan information to the server.

[0986] The server inputs the received plan information into a generative AI model. For example, an advanced artificial intelligence model such as OpenAI's GPT-4 is used as the generative AI model. The server converts the plan information into a prompt sentence and passes it to the generative AI model. An example of a prompt sentence is, "The user has entered a plan to distribute Ad A in the Tokyo area next Monday with an advertising budget of 50,000 yen. Please predict sales trends based on this plan."

[0987] The generative AI model predicts sales trends based on prompt statements. The model references past data and the results of similar advertising campaigns to generate predictions. For example, it might generate a prediction such as "Monday: 100,000 yen, Tuesday: 80,000 yen, Wednesday: 70,000 yen, Thursday: 60,000 yen, Friday: 50,000 yen."

[0988] The server receives the prediction results obtained from the generative AI model and sends them to the device, which then presents the results to the user, who can then view them through the device interface.

[0989] This system allows users to predict sales trends based on an advertising implementation plan and effectively utilize past advertising implementation results. This makes it easier to optimally allocate advertising expenses and plan effective advertising measures. The flow of the identification process in Example 3 will be described with reference to FIG. 15.

[0990] Step 1:

[0991] The user inputs an implementation plan for the advertising campaign. The user inputs a plan through the device interface, such as "Next Monday, in the Tokyo area, we will distribute Ad A with an advertising budget of 50,000 yen." The input plan information is temporarily stored in the device's memory.

[0992] Step 2:

[0993] The terminal sends the entered plan to the server. The terminal converts the plan information entered by the user into JSON format and sends it to the server using an HTTP request. If the transmission is successful, the terminal displays the message "Plan information sent." The input is the user's plan information, and the output is a transmission success message to the server.

[0994] Step 3:

[0995] The server inputs the plan information into the generative AI model. The server converts the received plan information into a prompt sentence. For example, it generates a prompt sentence such as, "The user has input a plan to distribute Ad A in the Tokyo area next Monday with an advertising budget of 50,000 yen. Please predict sales trends based on this plan." The input is the plan information, and the output is a prompt sentence.

[0996] Step 4:

[0997] The generative AI model predicts sales trends. The generative AI model predicts sales trends based on a prompt statement. The model references data on past advertising campaigns from an internal database to generate a predicted result. For example, the model might generate a predicted result such as "Monday: 100,000 yen, Tuesday: 80,000 yen, Wednesday: 70,000 yen, Thursday: 60,000 yen, Friday: 50,000 yen." The input is the prompt statement, and the output is the predicted sales trend.

[0998] Step 5:

[0999] The server receives the prediction results and sends them to the device. The server receives the prediction results obtained from the generative AI model, converts them to JSON format, and sends them to the device. If the transmission is successful, the server records a log stating "Prediction results sent." The input is the prediction results, and the output is a successful transmission message to the device.

[1000] Step 6:

[1001] The device presents the prediction results to the user. The device displays the prediction results received from the server on the screen. The user can check the prediction results on the device screen, such as "Monday: 100,000 yen, Tuesday: 80,000 yen, Wednesday: 70,000 yen, Thursday: 60,000 yen, Friday: 50,000 yen." The input is the prediction results, and the output is what is displayed to the user.

[1002] (Application example 3)

[1003] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1004] Conventional sales analysis systems for post-implementation internet advertising have had the problem of making it difficult to predict the effectiveness of advertising measures in advance, preventing advertisers from formulating optimal advertising strategies. Furthermore, there was also the problem of insufficient learning of sales trends based on past advertising results, resulting in low prediction accuracy.

[1005] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 3 is realized by the following means. In this invention, the server includes a means for a user to input an implementation plan for an advertising measure, a means including a generative AI model that predicts sales trends based on the implementation plan for the advertising measure, and a means for presenting the prediction results to the user. This allows advertisers to simulate the effects of their advertising measures in advance and develop optimal advertising strategies. In addition, the inclusion of a means for learning sales trends based on past advertising implementation results can improve prediction accuracy. Furthermore, the inclusion of an application to be installed on a smartphone allows users to easily check the predicted results of their advertising measures.

[1006] A "user" is an entity that inputs an implementation plan for advertising measures and checks the predicted results.

[1007] "Advertising implementation plan" refers to the specific details of the plan, such as the date, area, budget, and type of advertisement to be implemented.

[1008] A "generative AI model" is an artificial intelligence model that predicts sales trends based on the implementation plan of advertising campaigns entered by the user.

[1009] "Sales trends" is data showing fluctuations in sales after the implementation of advertising measures.

[1010] The "prediction result" is the forecast of future sales trends calculated by the generative AI model based on the implementation plan of the advertising campaign.

[1011] "Past advertising results" refers to the details of advertising measures previously implemented and the sales data obtained as a result.

[1012] A "smartphone" is a type of mobile phone, a multi-function device that can connect to the Internet and install applications.

[1013] An "application" is a software program that is installed on a smartphone and provides a specific function.

[1014] The system for implementing this invention uses a generative AI model that predicts sales trends based on an advertising implementation plan entered by the user. The system can be operated through an application installed on the user's smartphone.

[1015] The server receives the advertising implementation plan entered by the user and predicts sales trends based on that plan. Specifically, the server uses a generative AI model that learns sales trends based on past advertising results. This generative AI model is built using Python and its libraries (numpy, scikit-learn). The server calculates the prediction results and presents them to the user.

[1016] Users use an application installed on their smartphone to input plans for advertising initiatives. For example, they might input a plan such as "Next Monday, in the Tokyo area, we will distribute Ad A with an advertising budget of 50,000 yen." This input is done in text format.

[1017] An example of a prompt statement is as follows:

[1018] Please enter your advertising implementation plan.

[1019] Example: Next Monday, Ad A will be distributed in the Tokyo area with an advertising budget of 50,000 yen.

[1020] The server receives this input and uses the generative AI model to predict sales trends. The prediction results are displayed to users in real time via a smartphone application. This allows users to simulate the effects of their advertising campaigns in advance and develop optimal advertising strategies.

[1021] For example, if a user inputs a plan to "distribute Ad A next Monday in the Tokyo area with an advertising budget of 50,000 yen," the server will predict sales trends based on past data and present the results to the user as "Predicted sales: 15,000 yen." In this way, users can check the effectiveness of their advertising measures in advance and make appropriate decisions.

[1022] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[1023] Step 1:

[1024] The user launches the smartphone application and inputs the implementation plan for the advertising campaign, including the date, region, budget, and type of advertisement. The input data is sent to the server in text format.

[1025] Step 2:

[1026] The server receives the advertising plan sent by the user, analyzes the received data, and extracts necessary information (e.g., date, region, budget, and type of ad). Based on the analysis results, it prepares the data to be input into the generative AI model.

[1027] Step 3:

[1028] The server retrieves relevant data from a database of past advertising results. The retrieved data includes details of past advertising campaigns and the resulting sales data. This data is used to build a training dataset for the generative AI model.

[1029] Step 4:

[1030] The server uses a generative AI model to predict sales trends based on the implementation plan for advertising measures entered by the user. Specifically, it inputs new advertising measure data into a model trained on past data and predicts sales trends. The prediction results are calculated and output as numerical data.

[1031] Step 5:

[1032] The server sends the forecast results to the smartphone application to present to the user. The sent data is numerical data of the forecast sales.

[1033] Step 6:

[1034] The user's smartphone displays the received prediction results on the application screen, allowing the user to check the predicted sales results and simulate the effectiveness of advertising measures in advance.

[1035] Specifically, when a user inputs a plan such as "Distributing Ad A next Monday in the Tokyo area with an advertising budget of 50,000 yen," the server predicts sales trends based on past data and presents the results to the user as "Predicted sales: 15,000 yen." In this way, users can check the effectiveness of their advertising measures in advance and make appropriate decisions.

[1036] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1037] "Example 1"

[1038] In one embodiment of the present invention, the emotion engine recognizes the user's emotions. Specifically, it infers emotions from the user's facial expressions, tone of voice, and the language used in text messages. For example, when a user is happy, their voice tone tends to be higher and their language tends to be more positive. The AI ​​analyzes this information to recognize the user's emotional state.

[1039] "Example 2"

[1040] The emotion engine also evaluates the effectiveness of advertising campaigns based on the user's emotional state. Specifically, it analyzes how the user's emotional state changed after the advertising campaign was implemented. For example, if the user's emotional state clearly improved after the advertising campaign was implemented, the advertising campaign is evaluated as effective.

[1041] "Example 3"

[1042] Furthermore, the emotion engine suggests optimizing advertising measures based on the user's emotional state. Specifically, it learns patterns of advertising measures when the user's emotional state is good, and suggests new advertising measures based on those patterns. For example, if a user sees a specific ad A when they are happy, it will determine that ad A is likely to be well-received by the user and suggest a similar advertising measure.

[1043] The processing flow of each embodiment will be described below.

[1044] "Example 1"

[1045] Step 1: The emotion engine infers emotions from the user's facial expressions, tone of voice, and wording in text messages.

[1046] Step 2: AI analyzes this information and recognizes the user's emotional state.

[1047] "Example 2"

[1048] Step 1: After the advertising campaign is implemented, the emotion engine analyzes how the user's emotional state has changed.

[1049] Step 2: If the user's emotional state is clearly improved after the ad campaign is implemented, the ad campaign is evaluated as effective.

[1050] "Example 3"

[1051] Step 1: The emotion engine learns advertising patterns when the user is in a positive emotional state.

[1052] Step 2: Propose new advertising strategies based on those patterns.

[1053] Step 3: If a user sees a specific ad A while they are happy, we determine that ad A is likely to be well-received by the user and suggest a similar advertising strategy.

[1054] Example 1

[1055] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1056] Accurately analyzing the effectiveness of internet advertising and predicting future sales trends are important challenges for advertisers. However, conventional systems often require manual data collection, preprocessing, and output of prediction results, which is inefficient. Furthermore, they are unable to provide feedback that takes user emotions into account, making it difficult to use the data to improve advertising strategies. To solve these challenges, a system is needed that can automatically collect, preprocess, predict, and output data, as well as the ability to recognize and analyze user emotions.

[1057] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1058] In this invention, the server includes means for collecting sales data after the implementation of internet advertising, means for preprocessing the collected sales data, means for training an artificial intelligence model based on the preprocessed data, means for predicting future sales trends using the trained artificial intelligence model, means for outputting the prediction results in a format that is easy for the user to understand, means for recognizing user emotions, and means for analyzing the recognized emotion data. This enables automatic collection, preprocessing, prediction, and output of data, and further enables feedback that takes user emotions into consideration.

[1059] "Internet advertising" refers to advertising displayed on the Internet and delivered to users through websites, social media, search engines, etc.

[1060] "Sales data" refers to information regarding sales obtained after the implementation of advertising measures, including the number of clicks, conversion rate, sales amount, etc.

[1061] "Artificial intelligence" refers to technologies that allow computer systems to learn, reason, and self-correct by imitating human intelligence, and includes machine learning and deep learning.

[1062] "Preprocessing" refers to the process of converting collected data into a format suitable for analysis and learning, and includes filling in missing values, removing outliers, and normalizing data.

[1063] An "artificial intelligence model" is a mathematical model that is built to learn from data and perform specific tasks, including neural networks.

[1064] "Forecasting" means estimating future events and trends based on past data, and includes forecasting sales trends.

[1065] "Output" means that the system presents the processing results to the user, including display in the form of graphs or reports.

[1066] "Emotion recognition" is a technology that estimates a user's emotions from their facial expressions, tone of voice, and the wording of text messages, and it uses cameras, microphones, and natural language processing technology.

[1067] "Emotional data" refers to information about a user's emotional state, including emotional categories such as positive, negative, and neutral.

[1068] "Analysis" involves examining collected data in detail to find meaning and patterns, and includes processing to understand user feedback.

[1069] This invention is a system that includes artificial intelligence specialized in analyzing sales after the implementation of internet advertising. This system predicts future sales trends based on sales data collected after the implementation of advertising campaigns and presents them to users. It also has the ability to recognize user emotions and analyze feedback.

[1070] Hardware and software used

[1071] The server uses the following hardware and software:

[1072] Hardware: A server machine with a powerful processor, sufficient memory, and storage.

[1073] Software: Use APIs for data collection and machine learning frameworks such as TensorFlow and PyTorch for data preprocessing and training artificial intelligence models

[1074] The terminal uses the following hardware and software:

[1075] Hardware: A computer or smart device with a camera and microphone

[1076] Software: Emotion recognition software using natural language processing technology

[1077] Program processing explanation

[1078] The server collects sales data after internet advertising is carried out. Specifically, data is obtained via API from Google Analytics and advertising platforms (e.g., Google Ads, Facebook Ads). The collected data undergoes preprocessing, such as filling in missing values, removing outliers, and normalizing the data.

[1079] Based on the preprocessed data, the server trains an artificial intelligence model using TensorFlow or PyTorch. The trained model is then used to predict future sales trends. The prediction results are presented to the user in the form of a web dashboard or PDF report.

[1080] The device recognizes emotions from the user's facial expressions, tone of voice, and the wording of text messages using a camera, microphone, and natural language processing technology. The recognized emotion data is analyzed on the server and reflected as feedback to the user.

[1081] Specific examples

[1082] For example, consider the case where sales trends for November 2023 are predicted based on data from an online advertising campaign for a new product that was conducted in October 2023. The server collects data such as click counts, conversion rates, and sales amounts from Google Analytics and advertising platforms, and performs preprocessing. A neural network is built using TensorFlow, and a model is trained based on past advertising results and sales data. The trained model is used to predict sales for the next advertising campaign, and the results are displayed on a web dashboard.

[1083] Prompt Sentence Examples

[1084] "Based on data from an online advertising campaign for a new product in October 2023, please predict sales trends for November 2023. Display the results in a graph and analyze the user's emotional state."

[1085] In this way, the system not only analyzes the effectiveness of advertising campaigns and predicts future sales, but also recognizes user sentiment and provides feedback.

[1086] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1087] Step 1: Data collection

[1088] The server collects sales data after the execution of online advertising. Specifically, it obtains data from Google Analytics and advertising platforms (e.g., Google Ads, Facebook Ads) via API.

[1089] Input: Data such as clicks, conversion rate, and sales amount for your ad campaign

[1090] Output: Collected sales data

[1091] What happens: The server periodically sends API requests to retrieve data about advertising campaigns, which is then stored in a database.

[1092] Step 2: Data Preprocessing

[1093] The server preprocesses the collected sales data, specifically by filling in missing values, removing outliers, and normalizing the data.

[1094] Input: Collected sales data

[1095] Output: Preprocessed data

[1096] What it does: The server imputes missing values ​​with the average, removes abnormally high sales data, and standardizes the data to convert it into a format that is easy for the AI ​​model to learn.

[1097] Step 3: Training the AI ​​model

[1098] The server uses the preprocessed data to train an AI model, which uses frameworks such as TensorFlow and PyTorch.

[1099] Input: Preprocessed data

[1100] Output: A trained AI model

[1101] Specific operation: The server uses TensorFlow to build a neural network and trains the model based on past advertising results and sales data.

[1102] Step 4: Sales forecast

[1103] The server uses the trained AI model to predict future sales trends.

[1104] Input: Trained AI model, current advertising data

[1105] Output: Predicted future sales trends

[1106] What it does: The server predicts sales for the next advertising campaign and generates the results in a graph format.

[1107] Step 5: Output the results

[1108] The server outputs the prediction results in a user-friendly format, such as a web dashboard or PDF report.

[1109] Input: Forecast results of future sales trends

[1110] Output: Web dashboard, PDF report

[1111] What it does: The server displays the prediction results on a web dashboard and allows users to download them as a PDF report.

[1112] Step 6: Emotion Recognition

[1113] The device uses a camera, microphone, and natural language processing technology to recognize emotions from the user's facial expressions, tone of voice, and the wording of text messages.

[1114] Input: User's facial expressions, tone of voice, text messages

[1115] Output: Recognized emotion data

[1116] Specific operation: When a user accesses the system through a webcam, the device uses the camera and microphone to analyze the user's facial expressions and tone of voice to estimate their emotions.

[1117] Step 7: Analyze the sentiment data

[1118] The server analyzes the collected emotional data and recognizes the user's emotional state, which allows it to understand the user's feedback more effectively.

[1119] Input: Recognized emotion data

[1120] Output: Parsed emotional feedback

[1121] What it does: If a user gives positive feedback, the server will recognize that sentiment and incorporate it into the strategy for the next advertising campaign.

[1122] (Application example 1)

[1123] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1124] Conventional internet advertising effectiveness analysis systems can predict future sales trends based on sales data after the implementation of advertising measures, but they do not evaluate the effectiveness of advertising measures taking into account the emotional state of the user. This makes it difficult to propose optimal advertising measures to maximize advertising effectiveness.

[1125] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1126] In this invention, the server includes means for learning sales trends after the implementation of internet advertising and outputting predicted results of sales trends, means for presenting the predicted results to the user, means for analyzing the user's facial expressions and tone of voice to recognize the emotional state, and means for evaluating the effectiveness of advertising measures based on the emotional state. This makes it possible to evaluate the effectiveness of advertising measures taking the user's emotional state into consideration, and to propose optimal advertising measures to maximize the effectiveness of advertising.

[1127] "Internet advertising" refers to advertising displayed on the Internet, a means of providing information to users through websites and applications.

[1128] "Sales analysis" is the process of analyzing sales data obtained after implementing advertising measures and evaluating their effectiveness.

[1129] "AI" stands for artificial intelligence, a technology that learns from data and makes predictions and judgments.

[1130] "Sales trends" is data that indicates fluctuations in sales over a certain period of time.

[1131] The "predicted results" are the results of predicting future sales trends based on the data learned by the AI.

[1132] "User" means any individual or entity viewing an Internet advertisement.

[1133] "Facial expressions" refer to the movements of a user's face and changes in their facial muscles, and provide clues to their emotional state.

[1134] "Tone of voice" refers to vocal characteristics such as the pitch, strength, and intonation of a user's voice, which provide clues to their emotional state.

[1135] "Emotional state" refers to the user's psychological state, including emotions such as joy, sadness, and anger.

[1136] An "advertising program" is an advertising implementation method or strategy designed to achieve a specific objective.

[1137] "Effectiveness evaluation" is the process of evaluating the effectiveness of advertising measures.

[1138] The system for carrying out the present invention performs sales analysis taking into account the emotional state of the user in order to maximize the effectiveness of internet advertising. Specific embodiments are described below.

[1139] System Configuration

[1140] This system is configured using the following hardware and software.

[1141] Hardware: Smartphone camera, smartphone microphone

[1142] Software: OpenCV, TensorFlow / Keras, librosa

[1143] Program processing

[1144] 1. Facial Expression Recognition

[1145] The device (smartphone) captures the user's facial expressions using a camera, performs face detection using OpenCV, converts the detected facial region to grayscale, and then inputs it into an emotion recognition model, allowing the user's emotional state to be recognized in real time.

[1146] 2. Voice Emotion Recognition

[1147] The device records the user's voice using a microphone. Using librosa, the voice data is converted into MFCC features and input into a voice emotion recognition model. This allows the device to recognize the user's emotional state from the tone of their voice.

[1148] 3. Sales forecasting

[1149] The server collects and standardizes advertising campaign data, then inputs it into a sales forecasting model, which allows future sales trends to be predicted.

[1150] 4. Effectiveness evaluation

[1151] The server evaluates the effectiveness of advertising measures based on the user's emotional state and sales forecast results, allowing it to propose optimal advertising measures to maximize advertising effectiveness.

[1152] Specific examples

[1153] For example, if a user is viewing an advertisement on their smartphone and the camera captures their facial expressions, and the emotion recognition model detects "joy," the effectiveness of the advertising campaign can be evaluated based on that information, and future sales can be predicted.

[1154] Prompt Sentence Examples

[1155] When a user views an ad on their smartphone, the camera and microphone are used to analyze the user's facial expressions and tone of voice in real time, and an emotion engine is used to recognize the user's emotional state. Furthermore, sales data after viewing the ad is collected, and the AI ​​learns from this to predict future sales trends. This allows the system to propose optimal advertising measures to maximize the effectiveness of the ad.

[1156] In this way, it becomes possible to evaluate the effectiveness of advertising measures taking into account the emotional state of the user, and to propose optimal advertising measures to maximize the effectiveness of advertising.

[1157] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1158] Step 1:

[1159] The device captures the user's facial expressions using the smartphone camera. The input is the camera image, and the output is the captured facial image. Face detection is performed using OpenCV, and the detected facial area is converted to grayscale. This results in image data suitable for facial expression recognition.

[1160] Step 2:

[1161] The device inputs the grayscale facial image into the emotion recognition model. The input is a grayscale image, and the output is the emotion recognition result. The emotion recognition model is executed using TensorFlow / Keras to recognize the user's emotional state (e.g., joy, sadness, anger, etc.). This allows the user's emotional state to be grasped in real time.

[1162] Step 3:

[1163] The device records the user's voice using the smartphone's microphone. The input is audio data, and the output is a recorded audio file. librosa is used to convert the audio data into MFCC features, which converts the audio data into a format suitable for emotion recognition models.

[1164] Step 4:

[1165] The device inputs the voice data converted into MFCC features into a voice emotion recognition model. The input is the MFCC features, and the output is the voice emotion recognition result. The voice emotion recognition model is executed using TensorFlow / Keras, and the emotional state of the user is recognized from the tone of their voice. This makes it possible to understand the user's emotional state from their voice.

[1166] Step 5:

[1167] The server collects and standardizes advertising campaign data. The input is advertising campaign data, and the output is standardized data. StandardScaler is used to standardize the data and convert it into a format suitable for the sales forecasting model. This reduces data variability and improves model accuracy.

[1168] Step 6:

[1169] The server inputs the standardized advertising campaign data into the sales forecasting model. The input is the standardized data, and the output is the sales forecast result. The sales forecasting model is executed using TensorFlow / Keras to predict future sales trends. This allows the effectiveness of advertising campaigns to be evaluated numerically.

[1170] Step 7:

[1171] The server evaluates the effectiveness of advertising measures based on the user's emotional state and sales forecast results. The inputs are emotion recognition results and sales forecast results, and the output is the results of evaluating the effectiveness of advertising measures. This makes it possible to evaluate the effectiveness of advertising measures taking the user's emotional state into account, and to propose optimal advertising measures.

[1172] Example 2

[1173] Next, a description will be given of Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1174] Conventional internet advertising effectiveness analysis systems not only predict sales trends after the implementation of advertising measures, but also do not evaluate the effectiveness of advertising measures by taking into account the emotional state of the user. This makes it difficult to accurately evaluate the true effectiveness of advertising measures. In addition, data collection and preprocessing for evaluating the effectiveness of advertising measures are often done manually, resulting in inefficiencies.

[1175] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1176] In this invention, the server includes means for collecting past advertising implementation results and sales data, means for saving the collected data in a database, means for preprocessing the saved data, means for training a deep learning model using the preprocessed data, means for a user to input information about a new advertising campaign, means for predicting sales trends based on the input information, means for presenting the prediction results to the user, means for collecting the user's emotional state after the advertising campaign is implemented, means for sending the collected emotional data to an emotion engine, means for receiving analysis results from the emotion engine, means for evaluating the effectiveness of the advertising campaign based on the analysis results from the emotion engine, and means for presenting the evaluation results to the user. This makes it possible not only to predict sales trends after the implementation of advertising campaigns, but also to evaluate the effectiveness of advertising campaigns taking the user's emotional state into consideration.

[1177] "Past advertising results" refers to data on previously conducted advertising campaigns, including information such as the type of advertisement, the time of advertisement distribution, the distribution area, and the cost of advertisements.

[1178] "Sales data" refers to information relating to the sales quantity and sales amount of a product during a specific period.

[1179] A "database" is a system for efficiently storing, managing, and searching collected data.

[1180] "Preprocessing" refers to processes used to convert data into a format suitable for analysis and learning, such as filling in missing values, removing outliers, and normalizing data.

[1181] A "deep learning model" is an algorithm that uses a multi-layer neural network to learn the characteristics of data and perform predictions and classifications.

[1182] "User" means an individual or organization that uses the system to analyze and evaluate the effectiveness of advertising measures.

[1183] An "advertising initiative" is an advertising campaign or promotional activity designed to achieve a specific objective.

[1184] "Emotional state" refers to the emotional changes and psychological reactions that users experience after implementing advertising campaigns.

[1185] An "emotion engine" is software or a system for analyzing a user's emotional state and providing the results.

[1186] The "evaluation result" is the result of quantitatively or qualitatively evaluating the effectiveness of the advertising campaign.

[1187] This invention is a system that predicts and evaluates the effectiveness of advertising measures based on past advertising results and sales data. The system consists of three main elements: a server, a terminal, and a user.

[1188] Data collection and preprocessing

[1189] The server collects past advertising results and sales data. This data includes information such as the type of ad, ad delivery time, delivery area, and advertising cost. The collected data is stored in a MySQL database. The server then performs preprocessing on the data. Specifically, it performs tasks such as filling in missing values, removing outliers, and normalizing the data. For example, it removes data with abnormally high advertising costs and fills in missing sales data with the average value.

[1190] Data training

[1191] The server trains a deep learning model using the preprocessed data. This model is built using TensorFlow. The model architecture includes multiple layers (input layer, hidden layer, and output layer). The server trains the model using the type of ad, distribution time, distribution area, and advertising cost as input data and sales trends as output data. For example, the model can be trained using data from the past year.

[1192] Sales trend forecast

[1193] When a user wants to implement a new advertising campaign, they input the information from their device. For example, they input information about the advertising campaign, such as "broadcast a new TV commercial nationwide with advertising costs of 1 million yen." The server uses a deep learning model based on the input information to predict sales trends. For example, it generates a prediction that sales will increase by 20% in the next month. The prediction result is displayed on the device, and the user can check it.

[1194] Emotional state analysis

[1195] After the advertising campaign is implemented, the device conducts a survey of the user to collect the user's emotional state. For example, the user may input an emotional state such as "I felt good after seeing the advertisement." The server then sends the collected emotional data to the IBM Watson Tone Analyzer. The emotion engine analyzes the user's emotional state and sends the results to the server.

[1196] Evaluating the effectiveness of advertising measures

[1197] The server evaluates the effectiveness of the advertising campaign based on the analysis results of the emotion engine. For example, if the user's emotional state improves, the advertising campaign is evaluated as effective. The evaluation results are displayed on the device so the user can check them.

[1198] Examples of concrete examples and prompts

[1199] For example, if a user wants to implement an advertising campaign such as "broadcasting a new TV commercial nationwide with an advertising budget of 1 million yen," the user inputs this information into the device. Using a deep learning model, the server displays a predicted result that sales will increase by 20% in the next month. Furthermore, if the user responds after implementing the advertising campaign that "watching the ad made me feel good," the emotion engine will evaluate the advertising campaign as effective.

[1200] Example prompt sentence:

[1201] "We would like to implement a new advertising initiative. The type of ad will be a television commercial, aired during prime time, distributed nationwide, and cost 1 million yen. Please tell me your projected sales trends."

[1202] "Please tell us how the user's emotional state changed after implementing the advertising campaign. If the user's emotional state improved, please rate the advertising campaign as effective."

[1203] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1204] Step 1:

[1205] The server collects past advertising results and sales data. Specifically, it obtains data from marketing systems and sales management systems. As input, it receives information such as the type of advertisement, advertisement distribution time, distribution area, and advertising cost, and as output, it stores this data in a database.

[1206] Step 2:

[1207] The server stores the collected data in a database. Specifically, it stores the data in a MySQL database. It receives the data collected in step 1 as input and generates the data stored in the database as output.

[1208] Step 3:

[1209] The server preprocesses the stored data by imputing missing values, removing outliers, and normalizing the data. It receives raw data stored in the database as input and generates preprocessed data as output. For example, it removes data with abnormally high advertising costs and imputes missing sales data with the average value.

[1210] Step 4:

[1211] The server trains a deep learning model using the preprocessed data. Specifically, it uses TensorFlow to build a model and trains it using the training data as input. It receives the preprocessed data as input and generates a trained deep learning model as output.

[1212] Step 5:

[1213] The user inputs information about a new advertising campaign into the terminal. Specifically, the user inputs information such as the type of advertisement, distribution time, distribution area, and advertising cost. Detailed information about the advertising campaign is received as input, and this information is sent to the server as output.

[1214] Step 6:

[1215] The server predicts sales trends based on the input information. Specifically, it makes the prediction using a trained deep learning model. It receives advertising campaign information sent by the user as input and generates predicted sales trends as output. For example, it generates a prediction that sales will increase by 20% in the next month.

[1216] Step 7:

[1217] The server displays the prediction results on the terminal. Specifically, it presents the predicted sales trends to the user. It receives the prediction results as input and displays them on the terminal screen as output.

[1218] Step 8:

[1219] The user inputs their emotional state into the terminal after the implementation of the advertising campaign. Specifically, the emotional state is input in the form of a questionnaire. The emotional state after the implementation of the advertising campaign is received as input, and this information is sent to the server as output.

[1220] Step 9:

[1221] The server sends the collected emotion data to the emotion engine. Specifically, it sends the data to the emotion engine and requests it to analyze. It receives the emotion data sent by the user as input and sends it to the emotion engine as output.

[1222] Step 10:

[1223] The server receives the analysis results of the emotion engine. Specifically, the server receives the analysis results from the emotion engine. The server receives the analysis results from the emotion engine as input and stores the analysis results as output.

[1224] Step 11:

[1225] The server evaluates the effectiveness of the advertising campaign based on the analysis results of the emotion engine. Specifically, it evaluates whether the user's emotional state has improved. It receives the analysis results of the emotion engine as input and generates an evaluation result of the effectiveness of the advertising campaign as output.

[1226] Step 12:

[1227] The server displays the evaluation results on the terminal. Specifically, it presents the results of the evaluation of the effectiveness of the advertising campaign to the user. It receives the evaluation results as input and displays them on the terminal screen as output.

[1228] (Application example 2)

[1229] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1230] Conventional internet advertising effectiveness analysis systems are specialized in predicting sales trends, but do not take into account the emotional state of users when evaluating the effectiveness of advertising measures, making it difficult to accurately evaluate the true effectiveness of advertising. In addition, there is a lack of means to analyze user feedback text and calculate emotional scores, resulting in a lack of information needed to improve advertising measures. This makes it difficult for advertisers to plan and implement effective advertising measures.

[1231] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1232] In this invention, the server is a system including an AI specialized in analyzing sales after the implementation of internet advertising, and the AI ​​includes means for learning sales trends after the implementation of the campaign and outputting predicted results of the sales trends, means for presenting the predicted results to the user, means for evaluating the effectiveness of the advertising campaign based on the emotional state of the user, and means for analyzing the text of the user's feedback and calculating an emotional score. This makes it possible to simultaneously evaluate sales forecasts for advertising campaigns and the emotional state of the user.

[1233] "Internet advertising" refers to advertising delivered via the Internet and displayed to users through websites, social media, search engines, etc.

[1234] "Sales analysis" is the process of collecting and analyzing sales data over a specific period of time to identify sales trends and factors.

[1235] "AI" stands for artificial intelligence, a technology that allows machines to mimic human intelligence to learn, reason, and self-correct.

[1236] "Sales trends" are data that indicate fluctuations in sales over a certain period of time, and represent increases or decreases in sales over time.

[1237] The "estimated results" are the expected future sales figures that the AI ​​predicts based on past data.

[1238] "User" refers to consumers and advertisers who view internet advertisements.

[1239] "Emotional state" refers to the emotions a user feels when viewing an advertisement, including emotions such as joy, sadness, and excitement.

[1240] "Advertising measures" refer to advertising strategies and means planned and implemented to achieve a specific purpose.

[1241] "Evaluating effectiveness" means measuring and evaluating the extent to which advertising measures have achieved the desired results.

[1242] "Feedback" refers to opinions and feedback provided by users regarding advertisements.

[1243] "Text analysis" is the process of analyzing text data using natural language processing techniques to extract meaning and sentiment.

[1244] The "emotion score" is a numerical representation of a user's emotional state obtained through text analysis.

[1245] As an embodiment of the present invention, a system for analyzing the effectiveness of internet advertising and predicting sales trends is constructed. The system includes a server, a user terminal, and an AI model.

[1246] The server collects past advertising results and sales data, and uses this data to train a deep learning model. Specifically, it uses information such as ad type, distribution time, distribution area, and advertising cost as features to build a model that predicts sales trends. The server implements a random forest regression model using Python, and performs data preprocessing, model training, and predictions.

[1247] The user terminal provides an interface for advertisers to input information about new advertising initiatives. The user inputs information such as the type of advertisement, distribution time, distribution area, and advertising cost, and sends it to the server. The server uses this information to predict sales trends and returns the results to the user terminal.

[1248] Furthermore, the server collects user feedback and performs text analysis. Specifically, it uses the TextBlob library to analyze user feedback and calculate an emotion score. The emotion score is a numerical representation of the user's emotional state and is used to evaluate the effectiveness of advertising campaigns.

[1249] For example, if the type of advertisement is "video," the distribution time is "evening," the distribution area is "Tokyo," and the advertisement cost is "50,000 yen," the user enters the information as follows:

[1250] Ad Type: Video

[1251] Delivery time: Evening

[1252] Distribution area: Tokyo

[1253] Advertising cost: 50,000 yen

[1254] User Feedback: This ad made me so happy and excited!

[1255] The server uses this information to predict sales trends and analyzes user feedback to calculate an emotion score, allowing advertisers to evaluate the effectiveness of their advertising campaigns from both sales and emotion perspectives.

[1256] This system will provide advertisers with valuable information to plan and implement effective advertising strategies.

[1257] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1258] Step 1:

[1259] The server collects past advertising results and sales data. This includes information such as the type of ad, delivery time, delivery area, and advertising cost. This data is stored in a database for later processing. The input is past advertising data and sales data, and the output is the data stored in the database.

[1260] Step 2:

[1261] The server trains a deep learning model based on the collected data. Specifically, it implements a random forest regression model using Python and performs data preprocessing (such as imputing missing values ​​and normalizing the data). Next, it splits the data into training and test data and trains the model. The input is advertising data and sales data obtained from the database, and the output is the trained model.

[1262] Step 3:

[1263] The user terminal provides an interface for the advertiser to input information about new advertising initiatives. The user inputs information such as the type of advertisement, distribution time, distribution area, and advertising cost, and sends it to the server. The input is the advertising initiative information entered by the advertiser, and the output is the data sent to the server.

[1264] Step 4:

[1265] The server predicts sales trends based on the advertising campaign information sent by the user. Using the trained model, it predicts sales based on the features of the advertising campaigns entered. The input is the advertising campaign information sent by the user, and the output is predicted sales data.

[1266] Step 5:

[1267] The server returns the predicted sales data to the user terminal, which then displays the prediction results to the advertiser. The input is the predicted sales data, and the output is the prediction results displayed on the user terminal.

[1268] Step 6:

[1269] The user terminal collects user feedback on advertising campaigns. Users input their impressions and opinions on advertisements in text format and send them to the server. The input is the user feedback, and the output is the feedback data sent to the server.

[1270] Step 7:

[1271] The server analyzes the user's feedback text and calculates the emotion score. Specifically, it uses the TextBlob library to analyze the feedback and quantify the positive and negative emotions. The input is the user's feedback data, and the output is the calculated emotion score.

[1272] Step 8:

[1273] The server evaluates the effectiveness of the advertising campaign based on the calculated emotion score. It combines the sales forecast and emotion score to evaluate the overall effectiveness of the advertising campaign and returns the results to the user's device. The input is the sales forecast data and emotion score, and the output is the evaluation result of the advertising campaign's effectiveness.

[1274] Step 9:

[1275] The user terminal displays the results of the effectiveness evaluation of the advertising campaign to the advertiser. The advertiser uses this information to plan and implement their next advertising campaign. The input is the results of the effectiveness evaluation of the advertising campaign, and the output is the evaluation result displayed to the advertiser.

[1276] Example 3

[1277] Next, a description will be given of Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1278] Conventional post-sales analysis systems for internet advertising have had issues with accurately predicting the effectiveness of advertising measures and being unable to propose optimal advertising measures that take into account the emotional state of the user. This makes it difficult to find appropriate measures to maximize advertising effectiveness, which can result in wasted advertising costs and sluggish sales.

[1279] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[1280] In this invention, the server includes means for a user to input an implementation plan for an advertising campaign, means for a terminal to send the input information to the server, means for the server to preprocess the received information, means for the server to input a prompt sentence to a generative AI model and predict sales trends, means for the server to present the prediction results to the user, means for the server to collect the user's emotional state, and means for the server to propose optimization of the advertising campaign using an emotion engine. This makes it possible to accurately predict the effectiveness of advertising campaigns and further propose optimal advertising campaigns that take the user's emotional state into consideration.

[1281] A "user" is an entity that uses the system to input an implementation plan for advertising measures and receives forecast results of sales trends and proposals for optimizing advertising measures.

[1282] A "terminal" is a device that allows a user to input an implementation plan for an advertising campaign and transmit that information to a server.

[1283] The "server" is a computer system that receives information sent by users, performs preprocessing, and uses generative AI models and emotion engines to predict sales trends and propose optimized advertising strategies.

[1284] A "generative AI model" is an artificial intelligence model that predicts sales trends based on advertising campaign information entered by the user.

[1285] A "prompt" is an instruction entered into the generative AI model to predict sales trends.

[1286] The "emotion engine" is an artificial intelligence engine that analyzes the user's emotional state and uses that information to suggest ways to optimize advertising strategies.

[1287] The "advertising campaign implementation plan" is specific planning information entered by the user, such as the advertisement distribution date and time, region, advertising cost, and type of advertisement.

[1288] "Sales trends" is data indicating predicted fluctuations in sales after the implementation of advertising measures.

[1289] "Preprocessing" is data processing to convert information received from users into a format suitable for the generative AI model.

[1290] The "prediction result" is the predicted data of sales trends that the generative AI model outputs based on the prompt sentence.

[1291] "Optimization suggestions" are specific measures proposed by the emotion engine based on the user's emotional state to maximize the effectiveness of advertising campaigns.

[1292] MODE FOR CARRYING OUT THE INVENTION

[1293] This invention is a system that allows a user to input an implementation plan for an advertising campaign, predicts sales trends based on that information, and also proposes optimization of the advertising campaign taking into account the emotional state of the user. A specific embodiment of this system is described below.

[1294] System configuration

[1295] The system consists of a terminal where users can input their advertising implementation plans, a server that receives and processes the input information, a generative AI model that predicts sales trends, and an emotion engine that analyzes the user's emotional state.

[1296] Hardware and software used

[1297] Terminal: A device (e.g., PC, smartphone) through which a user inputs an advertising implementation plan.

[1298] Server: The computer system that receives information, performs preprocessing, and runs the generative AI model and emotion engine.

[1299] Generative AI models: Built using machine learning libraries such as TensorFlow and PyTorch.

[1300] Sentiment engine: Built using natural language processing models such as OpenAI's GPT-3 and BERT.

[1301] Data processing and calculation

[1302] 1. The user enters an advertising implementation plan: The user accesses the system's web interface and enters in a text box, "Next Monday, in the Tokyo area, we will distribute Ad A with an advertising budget of 50,000 yen."

[1303] 2. The terminal sends the entered information to the server: The terminal sends the information entered by the user to the server as an HTTP request. For example, it sends the information using a POST request.

[1304] 3. The server preprocesses the received information: The server parses the received information in JSON format, converts the date and time information into "YYYY-MM-DD" format, and converts the region information into a geographic code (e.g., Tokyo = 13).

[1305] 4. The server inputs a prompt into the generative AI model to predict sales trends: Based on the preprocessed information, the server generates a prompt such as "Please predict sales trends if Ad A is distributed in the Tokyo area next Monday with an advertising budget of 50,000 yen," and inputs it into the generative AI model. The generative AI model is built using TensorFlow.

[1306] 5. The server presents the prediction results to the user: The server generates the sales trend prediction results obtained from the generative AI model in HTML format and displays them in the user's browser. For example, it visually displays the predicted sales trend using a line graph.

[1307] 6. The server collects the user's emotional state: The server displays a questionnaire form asking the user about their emotional state. The user selects their emotional state from the options and clicks the submit button.

[1308] 7. The server uses an emotion engine to suggest optimization of advertising measures: The server uses an emotion engine (e.g., OpenAI's GPT-3) based on the collected emotion data to suggest optimization of advertising measures. For example, it generates a result such as "If a user sees a specific advertisement A when they are happy, it determines that advertisement A is likely to be well-received by the user, and suggests a similar advertising measure," and presents it to the user.

[1309] Examples of concrete examples and prompts

[1310] For example, a user might enter the following advertising plan:

[1311] "Next Monday, we will distribute Ad A in the Tokyo area with an advertising budget of 50,000 yen."

[1312] In this case, the server inputs the following prompt sentence into the generative AI model:

[1313] "Please predict the sales trend if Ad A is distributed in the Tokyo area next Monday with an advertising budget of 50,000 yen."

[1314] And the following prompt for the emotion engine:

[1315] "If a user sees a specific ad A when they are happy, determine that ad A is likely to be well-received by the user and suggest a similar advertising strategy."

[1316] In this way, the server provides the user with a sales trend forecast result and an optimized advertising measure. The flow of the identification process in the third embodiment will be described with reference to FIG.

[1317] Step 1:

[1318] The user inputs an implementation plan for the advertising campaign.

[1319] Input: The user enters details of the advertising campaign (e.g., "Next Monday, distribute Ad A in the Tokyo area with an advertising budget of 50,000 yen") into the system's web interface.

[1320] Output: Implementation plan data for the entered advertising campaign.

[1321] Specific operation: The user accesses the system's web interface, enters details of the advertising campaign in the text box, and clicks the submit button.

[1322] Step 2:

[1323] The terminal transmits the input information to the server.

[1324] Input: Advertising implementation plan data entered by the user.

[1325] Output: Advertising implementation plan data sent to the server.

[1326] Specific operation: The device sends the information entered by the user to the server as an HTTP POST request, which includes details of the advertising campaign.

[1327] Step 3:

[1328] The server preprocesses the information it receives.

[1329] Input: Advertising implementation plan data sent from the device.

[1330] Output: Preprocessed advertising campaign implementation plan data.

[1331] Specific operation: The server parses the received information in JSON format, converts the date and time information to "YYYY-MM-DD" format, converts the region information to a geographic code (e.g., Tokyo = 13), and checks for missing data and completes it if necessary.

[1332] Step 4:

[1333] The server inputs prompt statements into the generative AI model to predict sales trends.

[1334] Input: Preprocessed advertising campaign implementation plan data.

[1335] Output: Sales trend prediction results from the generative AI model.

[1336] Specific operation: Based on the preprocessed information, the server generates a prompt statement such as "Please predict the sales trend if Ad A is distributed in the Tokyo area next Monday with an advertising budget of 50,000 yen," and inputs it into the generative AI model. The generative AI model is built using TensorFlow.

[1337] Step 5:

[1338] The server presents the prediction results to the user.

[1339] Input: Sales trend prediction results obtained from the generative AI model.

[1340] Output: Sales forecast results presented to the user.

[1341] Specific operation: The server generates the sales trend prediction results obtained from the generative AI model in HTML format and displays them in the user's browser. For example, it visually displays the predicted sales trend using a line graph.

[1342] Step 6:

[1343] The server collects the user's emotional state.

[1344] Input: Data about the user's emotional state.

[1345] Output: Collected user emotional state data.

[1346] Specific operation: The server displays a questionnaire form asking the user about their emotional state. The user selects their emotional state from the options and clicks the submit button.

[1347] Step 7:

[1348] The server uses an emotion engine to propose optimization of advertising strategies.

[1349] Input: Collected user emotional state data.

[1350] Output: Optimization proposals for advertising measures using an emotion engine.

[1351] Specific operation: The server uses an emotion engine (e.g., OpenAI's GPT-3) based on the collected emotion data to propose optimization of advertising measures. For example, it generates a result such as "If a user sees a specific ad A when they are happy, it determines that ad A is likely to be well-received by the user, and suggests a similar advertising measure," and presents it to the user.

[1352] (Application example 3)

[1353] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1354] Conventional post-sales analysis systems for internet advertising have issues with predicting sales trends based on advertising implementation plans and not adequately optimizing advertising measures by taking into account the emotional state of users. This makes it difficult to maximize advertising effectiveness and may not meet advertisers' expectations.

[1355] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 3 is realized by the following means. In this invention, the server includes means for learning sales trends after the implementation of a campaign and outputting an estimated result of the sales trend, means for presenting the estimated result to the user, and means for analyzing the emotional state of the user and optimizing the advertising campaign. This makes it possible to predict sales trends based on an implementation plan of the advertising campaign and to optimize the advertising campaign taking into account the emotional state of the user.

[1356] "Internet advertising" refers to advertising delivered via the Internet.

[1357] "Sales analysis" refers to analyzing sales data obtained after implementing advertising measures and evaluating their effectiveness.

[1358] "AI" stands for artificial intelligence, a technology that mimics human intelligence through machine learning and data analysis.

[1359] "Sales trends" refers to data showing fluctuations in sales over a certain period of time.

[1360] "Predicted results" refers to data that shows future sales prospects predicted by AI.

[1361] "User" refers to advertisers and marketers who use the system.

[1362] "Emotional state" refers to the user's psychological state or feelings.

[1363] An "advertising strategy" is an advertising initiative designed to achieve a specific objective.

[1364] "Optimization" means adjusting conditions or parameters to obtain the best results in order to achieve a specific objective.

[1365] "Learning" is the process by which AI finds patterns and rules based on past data and uses them to make future predictions and decisions.

[1366] "Presentation" means that the system displays information to the user.

[1367] A system for carrying out this invention includes an AI specialized in analyzing sales after the implementation of internet advertising. The system includes means for learning sales trends after the implementation of a campaign and outputting predicted results of the sales trends, means for presenting the predicted results to a user, and means for analyzing the emotional state of the user and optimizing the advertising campaign.

[1368] 1. System Program

[1369] The system executes the following program:

[1370] Hardware: Smartphone

[1371] Software: Python, scikit-learn, sentiment analysis library (e.g., TextBlob), ad performance prediction library

[1372] 2. Program processing explanation

[1373] The server uses a sales forecasting model to predict sales trends based on the advertising campaign data entered by the user. Specifically, if a user enters a plan to "distribute Ad A in the Tokyo area next Monday with an advertising budget of 50,000 yen," the AI ​​will use this information to predict sales trends and present the results to the user.

[1374] Furthermore, the server uses an emotion engine to analyze the user's emotional state. For example, if the user is in a "happy" emotional state, the server optimizes advertising measures based on that emotional state. Specifically, if the user sees a specific advertisement A while feeling happy, the server determines that advertisement A is likely to be well-received by the user and suggests similar advertising measures.

[1375] 3. Examples and prompts

[1376] As a concrete example, consider the case where an advertiser inputs a plan to "distribute Ad A in the Tokyo area next Monday with an advertising budget of 50,000 yen." In this case, the system generates the following prompt sentence:

[1377] Example prompt sentence:

[1378] Enter your advertising strategy:

[1379] Region: Tokyo

[1380] Advertising cost: 50,000 yen

[1381] Ad Type: Ad A

[1382] Date: 2023-10-09

[1383] User's emotional state: joy

[1384] When this prompt sentence is input into the generative AI model, it outputs sales forecasts and optimized advertising measures, enabling advertisers to implement effective advertising measures.

[1385] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[1386] Step 1:

[1387] The user inputs the advertising plan, specifically the information such as region, advertising cost, advertising type, date, user's emotional state, etc. The input data is sent to the server.

[1388] input:

[1389] Region: Tokyo

[1390] Advertising cost: 50,000 yen

[1391] Ad Type: Ad A

[1392] Date: 2023-10-09

[1393] User's emotional state: joy

[1394] output:

[1395] Advertising campaign plan data sent to the server.

[1396] Step 2:

[1397] The server uses a sales forecasting model to predict sales trends based on the received advertising campaign plan data. Specifically, it uses past advertising execution results data and applies a machine learning algorithm (e.g., Linear Regression) to predict future sales trends.

[1398] input:

[1399] Advertising campaign planning data, past advertising results data.

[1400] output:

[1401] Projected sales trend data.

[1402] Step 3:

[1403] The server presents the predicted sales trend data to the user. Specifically, it generates sales trend graphs and numerical data and displays them on the user's terminal.

[1404] input:

[1405] Projected sales trend data.

[1406] output:

[1407] Sales trend graphs and numerical data displayed on the user's device.

[1408] Step 4:

[1409] The server analyzes the user's emotional state using an emotion engine. Specifically, it calculates an emotion score using an emotion analysis library (e.g., TextBlob) based on the emotional state data entered by the user.

[1410] input:

[1411] User emotional state data.

[1412] output:

[1413] The calculated sentiment score.

[1414] Step 5:

[1415] The server optimizes advertising measures based on the emotional score. Specifically, if the emotional score is high, it will suggest similar advertising measures, and if the emotional score is low, it will suggest changing the advertising type or region.

[1416] input:

[1417] Calculated sentiment scores and advertising strategy planning data.

[1418] output:

[1419] Proposal of optimized advertising measures.

[1420] Step 6:

[1421] The server presents the optimized advertising proposal to the user. Specifically, the proposal is generated in text format and displayed on the user's device.

[1422] input:

[1423] Proposal of optimized advertising measures.

[1424] output:

[1425] Proposing optimized advertising measures to be displayed on the user's device.

[1426] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1427] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1428] Another example of generative AI is Gemini (internet search engine). <url: https: gemini.google.com ?hl="ja">) are mentioned.

[1429] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[1430] [Third embodiment]

[1431] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1432] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[1433] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1434] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[1435] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1436] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1437] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1438] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1439] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1440] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1441] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1442] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[1443] "Example 1"

[1444] An example of the present invention is a system including AI specialized in analyzing sales after the implementation of internet advertising. This system collects sales data after the implementation of advertising measures, and the AI ​​learns from this data. The AI ​​predicts future sales trends based on past advertising results and sales data. The prediction results are output and presented to the user in a format that is easy for the user to understand.

[1445] "Example 2"

[1446] Specifically, AI uses deep learning models to learn and predict sales trends. Past advertising results and sales data include information such as the type of ad, ad delivery time, delivery area, and advertising costs. Based on this information, AI predicts sales trends after the implementation of advertising measures.

[1447] "Example 3"

[1448] More specifically, when a user inputs an implementation plan for an advertising campaign, the AI ​​predicts sales trends based on this plan. For example, if a user inputs a plan such as "Next Monday, in the Tokyo area, we will distribute Ad A with an advertising budget of 50,000 yen," the AI ​​will predict sales trends based on this information and present the results to the user.

[1449] The processing flow of each embodiment will be described below.

[1450] "Example 1"

[1451] Step 1: After the advertising campaign is implemented, the system collects sales data.

[1452] Step 2: The collected sales data is learned by AI.

[1453] Step 3: The AI ​​predicts future sales trends based on past advertising results and sales data.

[1454] Step 4: The prediction results are output in a user-friendly format and presented to the user.

[1455] "Example 2"

[1456] Step 1: The AI ​​uses a deep learning model to learn sales trends.

[1457] Step 2: Past advertising results and sales data includes information such as the type of ad, ad delivery time, delivery area, and advertising costs.

[1458] Step 3: Based on this information, the AI ​​predicts sales trends after the advertising campaign is implemented.

[1459] "Example 3"

[1460] Step 1: The user enters an implementation plan for their advertising campaign.

[1461] Step 2: The AI ​​predicts sales trends based on this plan.

[1462] Step 3: For example, if a user inputs a plan such as "Next Monday, in the Tokyo area, we will distribute Ad A with an advertising budget of 50,000 yen," the AI ​​will use this information to predict sales trends.

[1463] Step 4: The prediction results are presented to the user.

[1464] Example 1

[1465] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1466] Quantitatively evaluating the effectiveness of internet advertising and accurately predicting future sales trends is important for maximizing the effectiveness of advertising measures. However, with conventional methods, it was difficult to efficiently collect and analyze sales data after the implementation of advertising measures and predict future sales. Another challenge was presenting the prediction results in a format that was easy for users to understand.

[1467] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1468] In this invention, the server includes means for collecting sales data after the implementation of internet advertising, means for preprocessing the collected sales data, and means for training an artificial intelligence model using the preprocessed data, thereby making it possible to quantitatively evaluate the effectiveness of advertising measures and accurately predict future sales trends.

[1469] "Internet advertising" refers to advertising displayed on the Internet and delivered to users through websites, social media, search engines, etc.

[1470] "Sales data" refers to data relating to the sale of goods and services, and includes information such as the sales amount, sales quantity, and sales date and time.

[1471] "Artificial intelligence" refers to technologies that allow computer systems to learn, reason, and self-correct by imitating human intelligence, and includes machine learning and deep learning.

[1472] "Data collection means" refers to the methods and technologies for obtaining sales data after the implementation of internet advertising, including data acquisition through APIs.

[1473] "Data preprocessing means" refers to methods and techniques for preparing collected sales data in an analyzable format, including filling in missing values, removing outliers, and normalizing data.

[1474] An "artificial intelligence model" is a mathematical model that learns from data and makes predictions and classifications, and includes neural networks and regression models.

[1475] "Learning tools" refers to methods and techniques for training artificial intelligence models using pre-processed data, including machine learning algorithms and deep learning frameworks.

[1476] "Prediction means" refers to methods and technologies for predicting future sales trends using a trained artificial intelligence model, and includes the model's inference function.

[1477] "Presentation means" refers to the method or technique for visually displaying the prediction results to the user, including graphical and tabular presentations.

[1478] The present invention is a system including artificial intelligence specialized for analyzing sales after the execution of internet advertising. Specific embodiments of this system will be described below.

[1479] The server collects sales data after the online advertisements are executed. Specifically, the sales data is obtained through an API. For example, the sales data can be obtained using a web service API. The server uses the Python Requests library to send a request to the API endpoint and obtain the sales data.

[1480] Next, the server preprocesses the collected sales data, specifically by imputing missing values, removing outliers, normalizing the data, etc. The server uses the Pandas library to create a data frame, impute missing values, remove outliers, and normalize the data.

[1481] The server then uses the preprocessed data to train an artificial intelligence model. Specifically, it uses TensorFlow and PyTorch to build a model based on past advertising results and sales data. The server then uses TensorFlow to build a neural network model and trains it on the data.

[1482] After the learning is complete, the server uses the learned artificial intelligence model to predict future sales trends. Specifically, it predicts sales for the next three months. The server uses the trained model to predict sales for the next three months.

[1483] Finally, the terminal presents the prediction results received from the server to the user in the form of graphs and tables. The terminal uses the Matplotlib library to graph the prediction results and display them to the user.

[1484] As a concrete example, the following prompt sentence can be input to a generative AI model:

[1485] Example prompt sentence:

[1486] Collect sales data after implementing advertising campaign A and train the AI ​​model. Based on the advertising campaign and sales data from the past year, predict sales trends for the next three months.

[1487] In this way, users can quantitatively evaluate the effectiveness of their internet advertising and predict future sales. This system enables data-driven decision-making to maximize the effectiveness of advertising initiatives.

[1488] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1489] Step 1:

[1490] The server collects sales data after the implementation of online advertising. As input, it uses an API endpoint for obtaining sales data after the implementation of advertising campaigns. Specifically, the server uses the Python Requests library to send a request to the API and obtains the sales data in JSON format. The output is the obtained sales data.

[1491] Step 2:

[1492] The server preprocesses the collected sales data. It uses the sales data obtained in step 1 as input. Specifically, the server creates a data frame using the Pandas library, imputes missing values, removes outliers, and normalizes the data. The output is the preprocessed sales data.

[1493] Step 3:

[1494] The server uses the preprocessed data to train an artificial intelligence model. The sales data preprocessed in step 2 is used as input. Specifically, the server uses TensorFlow to build a neural network model and trains it on the data. The output is a trained artificial intelligence model.

[1495] Step 4:

[1496] The server uses the trained artificial intelligence model to predict future sales trends. As input, it uses the trained artificial intelligence model and data for the prediction period. Specifically, the server uses the trained model to predict sales for the next three months. The output is the predicted sales data.

[1497] Step 5:

[1498] The terminal presents the forecast results received from the server to the user. As input, it uses the sales data predicted in step 4. Specifically, the terminal uses the Matplotlib library to graph the forecast results and display them to the user. The output is the forecast results presented to the user.

[1499] (Application example 1)

[1500] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1501] Accurately analyzing the effectiveness of internet advertising and predicting future sales trends are important challenges for advertisers. However, conventional methods have made it difficult to analyze advertising campaign data in real time and forecast sales. There has also been a lack of ways to present the forecast results in a format that is easy for users to understand. This has made it difficult to make appropriate decisions to maximize the effectiveness of advertising measures.

[1502] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1503] In this invention, the server is a system including an AI specialized in analyzing sales after the implementation of internet advertising. The AI ​​includes means for learning sales trends after the implementation of a campaign and outputting predicted results of sales trends, means for presenting the predicted results to users, means for inputting advertising campaign data and making sales forecasts in real time, and means for displaying the predicted results in graphs and charts. This makes it possible to analyze advertising campaign data in real time and make sales forecasts. In addition, visually displaying the predicted results makes them easy for users to understand and supports appropriate decision-making.

[1504] "Internet advertising" refers to advertising displayed on the Internet, a means of providing information to users through websites and applications.

[1505] "Sales analysis" is the process of collecting and analyzing sales data over a specific period of time to identify sales trends and factors.

[1506] "AI" is an abbreviation for artificial intelligence, a technology that allows computers to imitate human intelligence to learn and reason.

[1507] "Sales trends after the implementation of the campaign" refers to the fluctuations in sales over a certain period of time after the implementation of the advertising campaign.

[1508] The "predicted results" are future sales forecasts based on the data learned by the AI.

[1509] "User" refers to any individual or legal entity that uses the System.

[1510] An "advertising campaign" is a planned advertising activity with a specific purpose, and may include multiple advertising initiatives.

[1511] "Real-time" refers to data being processed and analyzed as it is generated.

[1512] "Graphs and charts" are diagrams that visually represent numerical data, and include bar graphs, line graphs, pie charts, etc.

[1513] A "generative AI model" is an artificial intelligence model trained to perform a specific task, such as making predictions or classifications based on data.

[1514] A "prompt sentence" is a sentence that gives instructions or asks questions to the user, and serves to guide the operation of the system.

[1515] As an embodiment of the present invention, an advertising effectiveness analysis assistant system will be specifically described.

[1516] System configuration

[1517] This system includes AI specialized in analyzing sales after the execution of internet advertising. The system consists of the following main components:

[1518] 1. Data collection module: collects advertising campaign data and sends it to the server.

[1519] 2. Data Preprocessing Module: Cleans the collected data and converts it into a suitable format for analysis.

[1520] 3. AI model: Learns from past advertising results and sales data and predicts future sales trends.

[1521] 4. Prediction result display module: Visually displays the prediction results in graphs and charts.

[1522] Hardware and software used

[1523] Hardware: Smartphones, servers

[1524] Software: Python, pandas, scikit-learn, matplotlib

[1525] Data processing and calculation

[1526] The server processes the data in the following steps:

[1527] 1. Data Loading: The data collection module sends the advertising campaign data to the server as a CSV file, which the server loads.

[1528] 2. Data Preprocessing: The data preprocessing module removes missing values ​​and cleans the data.

[1529] 3. Model training: The AI ​​model trains a linear regression model targeting sales using ad spend, clicks, and impressions as features.

[1530] 4. Sales forecasting: The AI ​​model predicts sales based on new data.

[1531] 5. Displaying the results: The prediction result display module displays the prediction results as graphs and charts.

[1532] Specific examples

[1533] Users prepare their ad campaign data as a CSV file and upload it to the app. Then, they input the data for their new ad campaign (advertising costs, clicks, impressions). The server uses this data to generate sales forecasts and displays the results in graphs.

[1534] Prompt Sentence Examples

[1535] "Upload your ad campaign data and enter new campaign data. The AI ​​will generate sales forecasts and display the results in graphs."

[1536] In this way, the advertising effectiveness analysis assistant system can analyze advertising campaign data in real time and make sales forecasts. In addition, by visually displaying the forecast results, users can easily understand them and help them make appropriate decisions.

[1537] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1538] Step 1:

[1539] The user prepares advertising campaign data as a CSV file and uploads it to the app from their device. The input is the advertising campaign data, and the output is the CSV file sent to the server. Specifically, the user clicks the "Upload Data" button in the app and selects the CSV file from the file selection dialog.

[1540] Step 2:

[1541] The server reads the uploaded CSV file. The input is a CSV file, and the output is the advertising campaign data in a data frame. Specifically, the server converts the CSV file into a data frame using the Python pandas library.

[1542] Step 3:

[1543] The server uses a data preprocessing module to clean the data. The input is the advertising campaign data in a data frame format, and the output is the clean data with missing values ​​removed. Specifically, the server removes missing values ​​and converts the data into the required data type.

[1544] Step 4:

[1545] The server uses the AI ​​model to learn from past advertising results and sales data. The input is clean data, and the output is a trained AI model. Specifically, the server trains a linear regression model using the scikit-learn library.

[1546] Step 5:

[1547] The user inputs data for a new advertising campaign (advertising cost, clicks, impressions) into the app from their device. The input is the new advertising campaign data, and the output is the new data sent to the server. Specifically, the user enters the data into the app's input form and clicks the "Start prediction" button.

[1548] Step 6:

[1549] The server uses the trained AI model to make sales predictions based on new data. The input is new advertising campaign data, and the output is the sales prediction results. Specifically, the server inputs new data into the AI ​​model and obtains the prediction results.

[1550] Step 7:

[1551] The server uses the forecast result display module to visually display the forecast results in graphs and charts. The input is the sales forecast results, and the output is the graphs and charts presented to the user. Specifically, the server uses the matplotlib library to plot the forecast results on a graph and display it on the user's terminal.

[1552] The above are the specific processing steps of the advertising effectiveness analysis assistant system.

[1553] Example 2

[1554] Next, a description will be given of Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1555] Conventional internet advertising effectiveness analysis systems have difficulty accurately predicting sales trends after implementing advertising measures. In addition, there is a lack of effective ways to utilize past advertising results, and highly accurate predictions are required when planning advertising measures.

[1556] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1557] In this invention, the server includes means for collecting past advertising implementation results and sales data, means for preprocessing the collected data, means for training a deep learning model using the preprocessed data, means for a user to input information on new advertising measures, means for predicting sales trends based on the input information on the advertising measures, and means for presenting the prediction results to the user. This makes it possible to predict sales trends after the implementation of advertising measures with high accuracy and provide useful information to the user.

[1558] "Past advertising results" refers to detailed data when advertising was carried out, and includes information such as the type of advertisement, distribution time, distribution area, and advertising costs.

[1559] "Sales data" refers to data that indicates the sales performance of a product during a specific period, and includes information such as sales amount and sales quantity.

[1560] "Preprocessing" refers to the process of converting collected data into a format suitable for analysis and model training, and includes imputing missing values, normalizing data, encoding categorical data, etc.

[1561] A "deep learning model" is an algorithm that uses a multi-layer neural network to learn the characteristics of data and perform predictions and classifications.

[1562] "User" refers to an individual or corporation that uses the system to predict the effectiveness of advertising measures.

[1563] "Information about advertising measures" is detailed data about a new advertisement planned by the user, and includes information such as the type of advertisement, distribution time, distribution area, and advertising cost.

[1564] "Sales trend forecasting" refers to predicting fluctuations in sales after the implementation of advertising campaigns, and is done using a deep learning model.

[1565] The "prediction result" is a predicted value of sales trends calculated by a deep learning model and is presented to the user.

[1566] This invention is a system that predicts sales trends after implementing advertising measures based on past advertising results and sales data. The system consists of three main elements: a server, a terminal, and a user.

[1567] server

[1568] The server runs on the following hardware and software:

[1569] Hardware: A server machine equipped with a high-performance CPU, GPU, memory, and storage

[1570] Software: Deep learning frameworks such as TensorFlow and PyTorch, database management systems (e.g., MySQL, PostgreSQL)

[1571] The server first collects past advertising results and sales data, which involves retrieving data from the company's database or external APIs, and then stores the collected data in a database.

[1572] The server then preprocesses the collected data, specifically by imputing missing values, normalizing the data, encoding categorical data, etc. The preprocessed data is then converted into a format suitable for input into a deep learning model.

[1573] The server uses the preprocessed data to train a deep learning model. Training is performed using TensorFlow and PyTorch, and the model parameters are optimized. Once trained, the model is used to predict sales trends after the implementation of advertising campaigns.

[1574] Terminal

[1575] The terminal provides an interface for users to input information for new advertising campaigns. The terminal operates using the following hardware and software:

[1576] Hardware: PCs, tablets, smartphones, etc.

[1577] Software: Web browser, dedicated application

[1578] The user inputs information about the new advertising campaign into an input form on the device, including the type of advertisement, distribution time, distribution area, and advertising cost.

[1579] User

[1580] Users are individuals or corporations who use the system to predict the effectiveness of advertising campaigns. Users input information about new advertising campaigns through their terminals and receive prediction results.

[1581] Specific examples

[1582] For example, if a user inputs "A new TV ad will be broadcast in Tokyo from 8:00 PM to 10:00 PM, with an advertising cost of 1 million yen," the server inputs this information into the deep learning model and calculates the predicted sales trend. The server sends the calculation results to the terminal and displays to the user, "Predicted sales trend: 10% increase after one week, 5% increase after one month."

[1583] Prompt Sentence Examples

[1584] "Please predict the sales trend if a new TV ad is broadcast in Tokyo from 8:00 PM to 10:00 PM and the advertising cost is 1 million yen."

[1585] In this way, the server uses a deep learning model to predict the effectiveness of advertising campaigns and provide them to the user.

[1586] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1587] Step 1: Data collection

[1588] The server collects past advertising results and sales data from the company's database or external API. As input, it obtains information such as the type of ad, delivery time, delivery area, and advertising cost. As output, it stores this data in the database. Specifically, the server sends an API request and stores the obtained data in the database.

[1589] Step 2: Data Preprocessing

[1590] The server preprocesses the collected data. As input, it uses the advertising performance results and sales data stored in the database. As output, it generates the preprocessed data. Specifically, the server performs missing value imputation, data normalization, and categorical data encoding. For example, if there are missing values, it imputes them with the mean value, and standardizes numerical data.

[1591] Step 3: Model training

[1592] The server trains a deep learning model using the preprocessed data. It uses the preprocessed data as input and generates a trained deep learning model as output. Specifically, the server uses TensorFlow or PyTorch to build a neural network and optimizes the model parameters using the training data. For training, it uses the Adam optimizer or cross-entropy loss function.

[1593] Step 4: Enter a new advertising strategy

[1594] The user inputs information about a new advertising campaign into an input form on the device. As input, the user provides information such as the type of advertisement, distribution time, distribution area, and advertising cost. As output, the entered information about the advertising campaign is sent to the server. Specifically, the user inputs the information using a web browser or a dedicated application and clicks the send button.

[1595] Step 5: Prediction execution

[1596] The server inputs the information of new advertising campaigns entered by the user into the deep learning model and performs predictions. As input, it uses the information of advertising campaigns provided by the user. As output, it generates predicted sales trends. Specifically, the server inputs the information of advertising campaigns into the deep learning model, and the model calculates sales trends.

[1597] Step 6: View the results

[1598] The server sends the prediction results to the terminal and displays them to the user. The predicted sales trend is used as input. The prediction results are displayed to the user as output. Specifically, the server sends the prediction results to the terminal, and the terminal displays the results on the screen. For example, it displays results such as "Predicted sales trend: 10% increase after one week, 5% increase after one month."

[1599] (Application example 2)

[1600] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1601] Conventional internet advertising effectiveness analysis systems have difficulty accurately predicting sales trends after implementing advertising measures, and do not provide sufficient support for advertisers to develop optimal advertising strategies. Furthermore, there is no way for advertisers to easily predict the effectiveness of their advertising campaigns using smart devices. This has made it difficult for advertisers to optimally allocate advertising costs and implement effective advertising measures.

[1602] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1603] In this invention, the server includes means including an AI specialized in analyzing sales after the implementation of internet advertising, means for the AI ​​to learn sales trends after the implementation of a campaign and output predicted sales trends, means for presenting the predicted results to a user, and means for a user to predict the effectiveness of an advertising campaign using a smart device. This enables advertisers to learn sales trends based on past advertising results and predict sales trends based on information such as the type of advertisement, distribution time, distribution area, and advertising costs. In addition, advertisers can input data for a new advertising campaign using a smart device and predict sales, thereby receiving support for formulating an optimal advertising strategy.

[1604] "Internet advertising" refers to advertising distributed over the Internet, displayed to users through websites, social media, search engines, etc.

[1605] "Sales analysis" is the process of analyzing sales data obtained after implementing advertising measures and evaluating their effectiveness.

[1606] "AI" stands for artificial intelligence, which is a computer system designed to perform specific tasks automatically.

[1607] "Sales trends after the implementation of the campaign" refers to the fluctuations in sales over a certain period of time after the implementation of the advertising campaign.

[1608] The "predicted results" are future sales prospects predicted by AI based on past data.

[1609] "Users" refer to advertisers, marketers, and other people who use this system to predict advertising effectiveness.

[1610] A "smart device" is a portable electronic device with internet connectivity, such as a smartphone or tablet.

[1611] An "advertising campaign" is a series of advertising activities designed to achieve a specific objective.

[1612] "Ad type" refers to the format or media type of the ad, such as banner ad, video ad, or text ad.

[1613] "Serving time" refers to the time period during which an advertisement is displayed to a user.

[1614] "Distribution Area" means the geographic area in which an advertisement is displayed.

[1615] "Advertising costs" are expenses necessary to carry out an advertising campaign.

[1616] "Learning sales trends" is the process of building a model to predict future sales fluctuations based on past sales data.

[1617] "Predicting sales trends" means estimating future sales fluctuations based on past data.

[1618] An "advertising implementation plan" is a plan that includes the specific implementation details and schedule of an advertising campaign.

[1619] A system for carrying out the present invention includes an AI specialized for analyzing sales after the execution of internet advertising. A specific embodiment of this system will be described below.

[1620] System configuration

[1621] The system consists of the following main components:

[1622] 1. Server: Responsible for processing data and running AI models.

[1623] 2. Smart devices: Used by users to predict the effectiveness of advertising campaigns.

[1624] 3. Database: Stores past advertising results and sales data.

[1625] Hardware and software used

[1626] Hardware: Servers, smartphones, tablets

[1627] Software: Python, Pandas, Scikit-learn, TensorFlow / Keras

[1628] Data processing and calculation

[1629] The server processes the data in the following steps to predict sales trends.

[1630] 1. Data loading: The server loads past advertising results and sales data from the database.

[1631] 2. Data preprocessing: Using Pandas, categorical data such as ad type, delivery time, delivery area, and ad cost are converted into dummy variables, and the data is standardized using Scikit-learn.

[1632] 3. Building and training the model: Build a deep learning model using TensorFlow / Keras and train it with the training data.

[1633] 4. Sales forecasting: Input data for a new advertising campaign and use the AI ​​model to predict sales.

[1634] Specific examples

[1635] Forecast sales if an advertiser distributes a new banner ad in Tokyo at 12:00 and spends 50,000 yen on the ad.

[1636] Prompt Sentence Examples

[1637] What is the sales forecast for a new banner ad distributed in Tokyo at 12:00 with an advertising cost of 50,000 yen?

[1638] Processing flow

[1639] 1. User input: The user uses a smart device to input data for a new advertising campaign (ad type, delivery time, delivery area, advertising cost).

[1640] 2. Data transmission: Data is sent from the smart device to the server.

[1641] 3. Sales forecast: The server runs an AI model based on the received data to predict sales trends.

[1642] 4. Presenting the results: The server sends the prediction results to the smart device and presents them to the user.

[1643] In this way, advertisers can predict future sales based on past data and receive support in formulating optimal advertising strategies.

[1644] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1645] Step 1:

[1646] User Input

[1647] The user uses a smart device to enter data for a new advertising campaign (ad type, distribution time, distribution area, advertising cost), which is then formatted by the smart device application and prepared for transmission to the server.

[1648] Input: Ad type, delivery time, delivery area, ad cost

[1649] Output: Formatted ad campaign data

[1650] Step 2:

[1651] Sending data

[1652] The smart device sends the advertising campaign data entered by the user to a server, where it is securely transmitted over the internet.

[1653] Input: Formatted ad campaign data

[1654] Output: Ad campaign data sent to the server

[1655] Step 3:

[1656] Loading data

[1657] The server reads past advertising results and sales data from the database, including information on the type of ad, distribution time, distribution area, advertising cost, and sales.

[1658] Input: Past advertising results and sales data from the database

[1659] Output: Loaded past advertising results and sales data

[1660] Step 4:

[1661] Data Preprocessing

[1662] The server uses Pandas to convert categorical data such as ad type, delivery time, delivery area, and ad cost into dummy variables, and then uses Scikit-learn to standardize the data, preparing it in a format suitable for the AI ​​model.

[1663] Input: Loaded past advertising results and sales data

[1664] Output: Preprocessed data (dummy variables and standardized data)

[1665] Step 5:

[1666] Building and training the model

[1667] The server builds a deep learning model using TensorFlow / Keras and trains the model using the preprocessed data. The trained model has the ability to predict future sales based on past data.

[1668] Input: Preprocessed data

[1669] Output: A trained deep learning model

[1670] Step 6:

[1671] Sales forecast

[1672] The server preprocesses new ad campaign data submitted by users and uses the trained deep learning model to predict sales, outputting the results as a numerical value indicating the likelihood of future sales.

[1673] Input: New advertising campaign data, trained deep learning model

[1674] Output: Sales forecast results

[1675] Step 7:

[1676] Presentation of results

[1677] The server sends the prediction results to the smart device and displays them to the user, who can then check the results on the screen and use them to plan their advertising strategies.

[1678] Input: Sales Forecast Results

[1679] Output: Prediction results displayed on a smart device

[1680] In this way, advertisers can predict future sales based on past data and receive support in formulating optimal advertising strategies.

[1681] Example 3

[1682] Next, a third embodiment of the third embodiment will be described. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1683] With conventional sales analysis systems after the implementation of internet advertising, it was difficult to predict sales trends based on the implementation plan of advertising measures. In addition, there was a lack of means to effectively utilize past advertising results to predict sales trends, making it difficult to evaluate the effectiveness of advertising measures in advance. This made it difficult to optimally allocate advertising expenses and plan effective advertising measures.

[1684] The identification process by the identification processing unit 290 of the data processing device 12 in Example 3 is realized by the following means. In this invention, the server includes a means for a user to input an implementation plan for an advertising measure, a means for a terminal to transmit the input plan to the server, a means for the server to input plan information to the generative AI model, a means for the generative AI model to predict sales trends, a means for the server to receive and transmit the prediction results to the terminal, and a means for the terminal to present the prediction results to the user. This enables the user to predict sales trends based on the implementation plan for the advertising measure and effectively utilize past advertising implementation results.

[1685] A "user" is an entity that uses the system to input an implementation plan for advertising measures and receives forecast results for sales trends.

[1686] A "terminal" is a device through which a user inputs an implementation plan for an advertising campaign and communicates with a server to receive predicted results.

[1687] The "server" is a computer system that receives planning information sent by the user, inputs it into the generative AI model, generates prediction results, and sends them to the terminal.

[1688] A "generative AI model" is an artificial intelligence model that predicts sales trends based on an advertising implementation plan, and has the ability to learn from past advertising implementation results.

[1689] The "advertising measure implementation plan" is planning information including the details and conditions of the advertisement implementation input by the user.

[1690] "Sales trends" is data indicating predicted fluctuations in sales after the implementation of advertising measures.

[1691] "Prediction results" are sales trend forecast data calculated by the generative AI model based on the implementation plan of the advertising campaign.

[1692] This invention is a system that allows a user to input an implementation plan for an advertising campaign and predicts sales trends based on the plan. The system includes a user, a terminal, a server, and a generative AI model.

[1693] The user inputs an implementation plan for advertising measures through the device interface. For example, the user might input a plan such as "Next Monday, in the Tokyo area, we will distribute Ad A with an advertising budget of 50,000 yen." The device receives this input and sends the plan information to the server.

[1694] The server inputs the received plan information into a generative AI model. For example, an advanced artificial intelligence model such as OpenAI's GPT-4 is used as the generative AI model. The server converts the plan information into a prompt sentence and passes it to the generative AI model. An example of a prompt sentence is, "The user has entered a plan to distribute Ad A in the Tokyo area next Monday with an advertising budget of 50,000 yen. Please predict sales trends based on this plan."

[1695] The generative AI model predicts sales trends based on prompt statements. The model references past data and the results of similar advertising campaigns to generate predictions. For example, it might generate a prediction such as "Monday: 100,000 yen, Tuesday: 80,000 yen, Wednesday: 70,000 yen, Thursday: 60,000 yen, Friday: 50,000 yen."

[1696] The server receives the prediction results obtained from the generative AI model and sends them to the device, which then presents the results to the user, who can then view them through the device interface.

[1697] This system allows users to predict sales trends based on an advertising implementation plan and effectively utilize past advertising implementation results. This makes it easier to optimally allocate advertising expenses and plan effective advertising measures. The flow of the identification process in Example 3 will be described with reference to FIG. 15.

[1698] Step 1:

[1699] The user inputs an implementation plan for the advertising campaign. The user inputs a plan through the device interface, such as "Next Monday, in the Tokyo area, we will distribute Ad A with an advertising budget of 50,000 yen." The input plan information is temporarily stored in the device's memory.

[1700] Step 2:

[1701] The terminal sends the entered plan to the server. The terminal converts the plan information entered by the user into JSON format and sends it to the server using an HTTP request. If the transmission is successful, the terminal displays the message "Plan information sent." The input is the user's plan information, and the output is a transmission success message to the server.

[1702] Step 3:

[1703] The server inputs the plan information into the generative AI model. The server converts the received plan information into a prompt sentence. For example, it generates a prompt sentence such as, "The user has input a plan to distribute Ad A in the Tokyo area next Monday with an advertising budget of 50,000 yen. Please predict sales trends based on this plan." The input is the plan information, and the output is a prompt sentence.

[1704] Step 4:

[1705] The generative AI model predicts sales trends. The generative AI model predicts sales trends based on a prompt statement. The model references data on past advertising campaigns from an internal database to generate a predicted result. For example, the model might generate a predicted result such as "Monday: 100,000 yen, Tuesday: 80,000 yen, Wednesday: 70,000 yen, Thursday: 60,000 yen, Friday: 50,000 yen." The input is the prompt statement, and the output is the predicted sales trend.

[1706] Step 5:

[1707] The server receives the prediction results and sends them to the device. The server receives the prediction results obtained from the generative AI model, converts them to JSON format, and sends them to the device. If the transmission is successful, the server records a log stating "Prediction results sent." The input is the prediction results, and the output is a successful transmission message to the device.

[1708] Step 6:

[1709] The device presents the prediction results to the user. The device displays the prediction results received from the server on the screen. The user can check the prediction results on the device screen, such as "Monday: 100,000 yen, Tuesday: 80,000 yen, Wednesday: 70,000 yen, Thursday: 60,000 yen, Friday: 50,000 yen." The input is the prediction results, and the output is what is displayed to the user.

[1710] (Application example 3)

[1711] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1712] Conventional sales analysis systems for post-implementation internet advertising have had the problem of making it difficult to predict the effectiveness of advertising measures in advance, preventing advertisers from formulating optimal advertising strategies. Furthermore, there was also the problem of insufficient learning of sales trends based on past advertising results, resulting in low prediction accuracy.

[1713] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 3 is realized by the following means. In this invention, the server includes a means for a user to input an implementation plan for an advertising measure, a means including a generative AI model that predicts sales trends based on the implementation plan for the advertising measure, and a means for presenting the prediction results to the user. This allows advertisers to simulate the effects of their advertising measures in advance and develop optimal advertising strategies. In addition, the inclusion of a means for learning sales trends based on past advertising implementation results can improve prediction accuracy. Furthermore, the inclusion of an application to be installed on a smartphone allows users to easily check the predicted results of their advertising measures.

[1714] A "user" is an entity that inputs an implementation plan for advertising measures and checks the predicted results.

[1715] "Advertising implementation plan" refers to the specific details of the plan, such as the date, area, budget, and type of advertisement to be implemented.

[1716] A "generative AI model" is an artificial intelligence model that predicts sales trends based on the implementation plan of advertising campaigns entered by the user.

[1717] "Sales trends" is data showing fluctuations in sales after the implementation of advertising measures.

[1718] The "prediction result" is the forecast of future sales trends calculated by the generative AI model based on the implementation plan of the advertising campaign.

[1719] "Past advertising results" refers to the details of advertising measures previously implemented and the sales data obtained as a result.

[1720] A "smartphone" is a type of mobile phone, a multi-function device that can connect to the Internet and install applications.

[1721] An "application" is a software program that is installed on a smartphone and provides a specific function.

[1722] The system for implementing this invention uses a generative AI model that predicts sales trends based on an advertising implementation plan entered by the user. The system can be operated through an application installed on the user's smartphone.

[1723] The server receives the advertising implementation plan entered by the user and predicts sales trends based on that plan. Specifically, the server uses a generative AI model that learns sales trends based on past advertising results. This generative AI model is built using Python and its libraries (numpy, scikit-learn). The server calculates the prediction results and presents them to the user.

[1724] Users use an application installed on their smartphone to input plans for advertising initiatives. For example, they might input a plan such as "Next Monday, in the Tokyo area, we will distribute Ad A with an advertising budget of 50,000 yen." This input is done in text format.

[1725] An example of a prompt statement is as follows:

[1726] Please enter your advertising implementation plan.

[1727] Example: Next Monday, Ad A will be distributed in the Tokyo area with an advertising budget of 50,000 yen.

[1728] The server receives this input and uses the generative AI model to predict sales trends. The prediction results are displayed to users in real time via a smartphone application. This allows users to simulate the effects of their advertising campaigns in advance and develop optimal advertising strategies.

[1729] For example, if a user inputs a plan to "distribute Ad A next Monday in the Tokyo area with an advertising budget of 50,000 yen," the server will predict sales trends based on past data and present the results to the user as "Predicted sales: 15,000 yen." In this way, users can check the effectiveness of their advertising measures in advance and make appropriate decisions.

[1730] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[1731] Step 1:

[1732] The user launches the smartphone application and inputs the implementation plan for the advertising campaign, including the date, region, budget, and type of advertisement. The input data is sent to the server in text format.

[1733] Step 2:

[1734] The server receives the advertising plan sent by the user, analyzes the received data, and extracts necessary information (e.g., date, region, budget, and type of ad). Based on the analysis results, it prepares the data to be input into the generative AI model.

[1735] Step 3:

[1736] The server retrieves relevant data from a database of past advertising results. The retrieved data includes details of past advertising campaigns and the resulting sales data. This data is used to build a training dataset for the generative AI model.

[1737] Step 4:

[1738] The server uses a generative AI model to predict sales trends based on the implementation plan for advertising measures entered by the user. Specifically, it inputs new advertising measure data into a model trained on past data and predicts sales trends. The prediction results are calculated and output as numerical data.

[1739] Step 5:

[1740] The server sends the forecast results to the smartphone application to present to the user. The sent data is numerical data of the forecast sales.

[1741] Step 6:

[1742] The user's smartphone displays the received prediction results on the application screen, allowing the user to check the predicted sales results and simulate the effectiveness of advertising measures in advance.

[1743] Specifically, when a user inputs a plan such as "Distributing Ad A next Monday in the Tokyo area with an advertising budget of 50,000 yen," the server predicts sales trends based on past data and presents the results to the user as "Predicted sales: 15,000 yen." In this way, users can check the effectiveness of their advertising measures in advance and make appropriate decisions.

[1744] Furthermore, an emotion engine that estimates the user's emotion may be combined.

[1745] The identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1746] "Example 1"

[1747] In one embodiment of the present invention, the emotion engine recognizes the user's emotions. Specifically, it infers emotions from the user's facial expressions, tone of voice, and the language used in text messages. For example, when a user is happy, their voice tone tends to be higher and their language tends to be more positive. The AI ​​analyzes this information to recognize the user's emotional state.

[1748] "Example 2"

[1749] The emotion engine also evaluates the effectiveness of advertising campaigns based on the user's emotional state. Specifically, it analyzes how the user's emotional state changed after the advertising campaign was implemented. For example, if the user's emotional state clearly improved after the advertising campaign was implemented, the advertising campaign is evaluated as effective.

[1750] "Example 3"

[1751] Furthermore, the emotion engine suggests optimizing advertising measures based on the user's emotional state. Specifically, it learns patterns of advertising measures when the user's emotional state is good, and suggests new advertising measures based on those patterns. For example, if a user sees a specific ad A when they are happy, it will determine that ad A is likely to be well-received by the user and suggest a similar advertising measure.

[1752] The processing flow of each embodiment will be described below.

[1753] "Example 1"

[1754] Step 1: The emotion engine infers emotions from the user's facial expressions, tone of voice, and wording in text messages.

[1755] Step 2: AI analyzes this information and recognizes the user's emotional state.

[1756] "Example 2"

[1757] Step 1: After the advertising campaign is implemented, the emotion engine analyzes how the user's emotional state has changed.

[1758] Step 2: If the user's emotional state is clearly improved after the ad campaign is implemented, the ad campaign is evaluated as effective.

[1759] "Example 3"

[1760] Step 1: The emotion engine learns advertising patterns when the user is in a positive emotional state.

[1761] Step 2: Propose new advertising strategies based on those patterns.

[1762] Step 3: If a user sees a specific ad A while they are happy, we determine that ad A is likely to be well-received by the user and suggest a similar advertising strategy.

[1763] Example 1

[1764] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1765] Accurately analyzing the effectiveness of internet advertising and predicting future sales trends are important challenges for advertisers. However, conventional systems often require manual data collection, preprocessing, and output of prediction results, which is inefficient. Furthermore, they are unable to...

Claims

1. A system including an AI specialized in sales analysis after the implementation of internet advertising, A means for the AI ​​to learn sales trends after the implementation of internet advertising and output a predicted result of sales trends based on an implementation plan of an advertising campaign; A means for presenting the estimation result to a first user who intends to implement the advertising measure; A means for analyzing an emotional state of a second user who viewed an Internet advertisement and optimizing the advertising measure; A system including:

2. A means for the AI ​​to learn sales trends based on past advertising results; A means for proposing an advertising measure based on the emotional state; The system of claim 1 , comprising:

3. The means for outputting the inference result outputs the inference result generated using the generative AI model based on a prompt sentence including information on an implementation plan of the advertising campaign. The system of claim 1 .

Images