System

Abstract

A system is provided.SOLUTION: A system comprising: means for collecting historical sales data, external market trend data, and seasonal data; means for cleaning and pre-processing the collected data; means for training a machine learning model for demand forecasting using the pre-processed data; means for forecasting future sales trends and demand using the trained model; means for generating a dashboard for displaying the forecast results; means for dynamic pricing based on the forecast results; and means for generating individually optimized marketing strategies based on customer data.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] Conventional demand forecasting methods rely on experience and intuition, which limits their accuracy and results in problems such as not optimizing a company's sales plans and inventory management. Furthermore, it is difficult to generate dynamic pricing and personalized marketing strategies, which hinders profit maximization. The purpose of this invention is to resolve these problems and enable highly accurate demand forecasting, dynamic pricing, and individually optimized marketing strategies, thereby optimizing a company's sales plans and inventory management. [Means for solving the problem]

[0005] The present invention solves the above-mentioned problems with a system including: means for collecting historical sales data, external market trend data, and seasonal data; means for cleaning and preprocessing the collected data; means for training a machine learning model for demand forecasting using the preprocessed data; means for forecasting future sales trends and demand using the trained model; means for generating a dashboard for displaying the forecast results; means for dynamic pricing based on the forecast results; and means for generating individually optimized marketing strategies based on customer data. Specifically, the system improves the accuracy of demand forecasting by imputing missing values ​​and removing outliers based on the collected data, dividing the preprocessed data into training data and validation data, and training a time series forecasting model or a neural network model using the training data. Furthermore, by dynamic pricing based on the forecast results and generating individually optimized marketing strategies, the system can optimize a company's sales planning, inventory management, and marketing.

[0006] "Sales data" refers to data relating to the quantity, price, date and time of products actually sold by a company in the past.

[0007] "Market trend data" is data that indicates trends in consumer interest and demand in a particular market, often collected from external data sources.

[0008] "Seasonal data" refers to data that indicates sales fluctuations associated with particular seasons or events, including calendar events and other factors.

[0009] "Data cleaning" is the process of detecting missing or outliers from collected data and completing or removing them.

[0010] "Preprocessing" refers to a series of steps that transform data into a form that is easy for machine learning models to handle, including scaling, normalization, and integration.

[0011] A "machine learning model" is an algorithm used to predict future demand and sales trends using collected and pre-processed data.

[0012] "Training" is the process of repeatedly feeding data into a machine learning model so that the model learns patterns.

[0013] A "dashboard" is a screen or interface for visually displaying forecast results, including graphs and charts.

[0014] "Dynamic pricing" refers to adjusting product prices in real time based on demand forecasts.

[0015] A "marketing strategy" is a method for planning and implementing advertising and promotional activities aimed at specific customer segments. [Brief explanation of the drawings]

[0016] [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. 3 is a sequence diagram illustrating a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0017] 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.

[0018] First, the terms used in the following description will be explained.

[0019] 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, a 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), and an APU (Accelerated Processing Unit).

[0020] 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.

[0021] 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.

[0022] 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.

[0023] 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."

[0024] [First embodiment]

[0025] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0026] 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.

[0027] 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).

[0028] 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.

[0029] 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.

[0030] 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.

[0031] 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.

[0032] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0033] 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.

[0034] 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.

[0035] 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.

[0036] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. 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."

[0037] As an embodiment of the present invention, we provide a system that uses AI and machine learning to forecast future sales trends and demand. This system aims to enable companies to effectively optimize sales planning, inventory management, and marketing strategies by using past sales data, market trend data, and seasonal data.

[0038] The specific program processing of this system is as follows.

[0039] First, the server accesses the company's database to collect past sales data, external market trend data, and seasonal data. For example, the server obtains monthly sales data for the past five years and then obtains market trend data from an external API. In this way, multifaceted data collection forms the basis for a more accurate forecasting model.

[0040] The server then cleans and preprocesses the collected data. For example, missing values ​​in the sales data are imputed using the mean or median. Outliers are also removed. This preprocessing phase normalizes the data and integrates market trend and seasonal data with the sales data. The resulting dataset is then used to train the machine learning model.

[0041] The server then uses the preprocessed data to train a machine learning model. Here, the data is split into training data and validation data, and a time series forecasting model or neural network model is selected and trained. For example, the server uses an LSTM (long short-term memory) model to train a model to predict sales demand for the next month. During this training process, the data is repeatedly fed into the model, and patterns are learned.

[0042] Once training is complete, the server uses the newly collected data to predict future sales trends and demand. These predictions are stored locally and then made available to users via a dashboard. For example, the dashboard might display sales forecasts for a specific product during the upcoming Christmas season.

[0043] Once the forecast results are available, the server then performs dynamic pricing, adjusting product prices in real time based on predicted demand. For example, discounts are applied to items with high inventory, and prices are increased for items predicted to be in high demand. This dynamic pricing allows businesses to maximize profits.

[0044] Furthermore, the user implements individually optimized marketing strategies provided by the server. The server generates personalized promotional emails for specific customer segments based on the prediction results and customer data. The user can then run advertising campaigns based on this and monitor the results in real time.

[0045] In this way, the present invention provides dynamic pricing and optimized marketing strategies based on highly accurate demand forecasts, realizing a system that contributes to corporate sales planning, inventory management, and profit maximization.

[0046] The processing flow will be explained below.

[0047] Step 1:

[0048] The server accesses the company's database and collects past sales data, such as monthly sales data for the past five years and detailed information such as the sales volume and price of each product.

[0049] Step 2:

[0050] The server uses external APIs to collect relevant market trend data, specifically Google Trends and social media data, to understand consumer interest trends in the market.

[0051] Step 3:

[0052] The server accesses a calendar event database to retrieve data that reflects seasonality, for example, collecting data about seasonal events such as Christmas and Valentine's Day.

[0053] Step 4:

[0054] The server cleans the collected data by detecting missing values ​​in the dataset, imputing them with the mean or median, and removing outliers.

[0055] Step 5:

[0056] The server pre-processes the cleaned data, including scaling and normalizing the data, and integrating sales data, market trend data, and seasonality data.

[0057] Step 6:

[0058] The server splits the preprocessed dataset into training and validation data, typically 80% of the total data for training and 20% for validation.

[0059] Step 7:

[0060] The server selects a time series prediction model or neural network model, for example, a long short-term memory (LSTM) model.

[0061] Step 8:

[0062] The server uses the training data to train the machine learning model, which involves repeatedly feeding the data into the model and letting it learn patterns.

[0063] Step 9:

[0064] The server evaluates the performance of the trained model using validation data, using metrics such as MSE (Mean Squared Error) and MAE (Mean Absolute Error).

[0065] Step 10:

[0066] The server uses the new data to predict future sales trends and demand, for example, predicting the sales volume of a particular product for the next month.

[0067] Step 11:

[0068] The server stores the prediction results in a database and generates a dashboard, which displays the prediction results in the form of graphs and charts.

[0069] Step 12:

[0070] The device displays the prediction results to the user through a dashboard, providing visual information that is easy for the user to understand.

[0071] Step 13:

[0072] The server dynamically sets prices based on the demand forecast results, for example, raising prices for products with high predicted demand and discounting products with high inventory.

[0073] Step 14:

[0074] The user implements individually optimized marketing strategies generated by the server, for example, sending personalized promotional emails to specific customer segments.

[0075] Step 15:

[0076] Users can monitor the effectiveness of their marketing efforts in real time, allowing them to quickly determine the success of their efforts and make any necessary adjustments.

[0077] Example 1

[0078] Next, a description will be given of 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."

[0079] Conventional sales forecasting systems often base their forecasts solely on past sales data and are unable to comprehensively incorporate external market trend data or seasonal data, resulting in low forecast accuracy. Furthermore, they lack effective means for implementing dynamic pricing and individually optimized marketing strategies. A system that can solve this issue and provide highly accurate demand forecasts and dynamic pricing and marketing strategies based on those forecasts is needed.

[0080] 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.

[0081] In this invention, the server includes means for collecting past sales data, external market trend data, and seasonal data, means for cleaning and pre-processing the collected data, means for training a machine learning model for demand forecasting using the pre-processed data, means for forecasting future sales trends and demand using the trained model, means for generating a recording medium for displaying the forecast results, means for dynamic pricing based on the forecast results, and means for generating an individually optimized marketing strategy based on customer data, thereby enabling highly accurate demand forecasting, dynamic pricing, and the implementation of individually optimized marketing strategies.

[0082] "Sales data" is a collection of information about sales performance of a particular product over a past period.

[0083] "Market trend data" is a collection of information that indicates the current situation and future trends in a particular market or industry.

[0084] "Seasonal data" is a collection of information that indicates sales trends and demand fluctuations associated with particular seasons or annual events.

[0085] "Data cleaning" is the process of removing inaccurate, incomplete, or unnecessary values ​​from collected data to improve data quality.

[0086] "Preprocessing" refers to a series of steps taken to convert collected data into a format suitable for analysis and machine learning.

[0087] A "machine learning model" is an algorithm or mathematical model that learns patterns from data and performs tasks such as prediction and classification.

[0088] "Training" is the process of providing a machine learning model with large amounts of data and having it learn from that data to improve its prediction accuracy.

[0089] A "recording medium" is a physical or electronic device or means for storing and displaying digital data.

[0090] "Dynamic pricing" is a method of adjusting prices in real time based on demand forecasts and market conditions.

[0091] A "marketing strategy" is a series of plans and measures implemented to promote sales and increase brand awareness among specific customer segments.

[0092] This invention relates to a sales forecasting system, specifically a system that uses AI and machine learning technologies to forecast future sales trends and demand. This system is designed to help companies more effectively plan and execute sales plans, inventory management, and marketing strategies by utilizing past sales data, market trend data, and seasonal data.

[0093] Hardware and software used

[0094] 1. Server: Manages all data collection, preprocessing, machine learning model training, and dynamic pricing using libraries such as Python, Pandas, Scikit-learn, TensorFlow, or PyTorch.

[0095] 2. Database: Stores and manages company sales data and external trend data. Specifically, MySQL or PostgreSQL can be used.

[0096] 3. External API: Used to obtain market trend data.

[0097] 4. Web dashboard: An interface for providing predictions and analysis results to users. Specifically, JavaScript, D3.js, or Tableau may be used.

[0098] Data collection and preprocessing

[0099] The server accesses the company's database to collect past sales data, external market trend data, and seasonal data. This data allows for more accurate forecasts based on multifaceted information. Specifically, the server periodically sends API requests to collect and store the necessary data.

[0100] Next, the server cleans and preprocesses the data. This step includes imputing missing values, removing outliers, and normalizing the data. For example, it uses the Pandas fillna() method to impute missing values ​​with the mean.

[0101] Training a machine learning model

[0102] The server uses the preprocessed data to train a machine learning model. Here, the data is divided into training data and validation data, and a long short-term memory (LSTM) model is used to perform time series prediction. Specifically, an LSTM model is built and trained using TensorFlow.

[0103] Future sales trends and demand forecasts

[0104] Using the trained model, the server predicts future sales trends and demand based on newly collected data. These predictions are stored in a database and made available to users through a web dashboard. For example, the dashboard might predict that demand for a particular product will increase during the upcoming Christmas season.

[0105] Dynamic Pricing

[0106] The server then dynamically sets prices based on the forecast, for example, discounting items with high inventory and raising prices for items predicted to be in high demand. This dynamic pricing is a key factor in helping companies maximize profits.

[0107] Individually optimized marketing strategies

[0108] Based on the prediction results and customer data provided by the server, users can create personalized promotional emails and advertising campaigns for specific customer segments. For example, emails can be automatically generated to strongly appeal to specific customer demographics about a specific product.

[0109] Prompt Sentence Examples

[0110] "Please provide us with your sales forecast for electronic products for the upcoming Christmas season. Also, please identify which products will sell best and suggest a corresponding marketing strategy."

[0111] In this way, this system provides dynamic pricing and optimized marketing strategies based on highly accurate demand forecasts, contributing to corporate sales planning, inventory management, and profit maximization.

[0112] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0113] Step 1:

[0114] Data collection

[0115] The server accesses the company's database to collect historical sales data, external market trend data, and seasonality data.

[0116] Specific behavior:

[0117] The server periodically executes SQL queries to extract sales data from the company's database. The server sends HTTP requests to retrieve market trend data from an external API. The server stores this data in a local database.

[0118] input:

[0119] Corporate database connection information, external API endpoints.

[0120] output:

[0121] Collection of historical sales data, market trend data, and seasonality data.

[0122] Step 2:

[0123] Data Cleaning and Preprocessing

[0124] The server cleans the collected data, filling in missing values, removing outliers, and normalizing it.

[0125] Specific behavior:

[0126] The server uses the Pandas library to load the data into a data frame, uses the fillna() method to impute missing values, boxplots and Z-scores to detect and remove outliers, and normalizes each data set to create a unified dataset.

[0127] input:

[0128] Collected sales data, market trend data, and seasonality data.

[0129] output:

[0130] Cleaned and preprocessed integrated dataset.

[0131] Step 3:

[0132] Data partitioning

[0133] The server splits the combined dataset into training data and validation data.

[0134] Specific behavior:

[0135] The server splits the data into training and validation data using Scikit-learn's train_test_split function.

[0136] input:

[0137] Cleaned and preprocessed integrated dataset.

[0138] output:

[0139] Training and validation data.

[0140] Step 4:

[0141] Training a machine learning model

[0142] The server trains the LSTM model using the training data.

[0143] Specific behavior:

[0144] The server uses the TensorFlow library to build an LSTM model and trains it using the model.fit() method. During the training process, the validation data is used to evaluate the accuracy of the model.

[0145] input:

[0146] Training data, validation data.

[0147] output:

[0148] A trained LSTM model.

[0149] Step 5:

[0150] Future predictions

[0151] The server uses the trained model to predict future sales trends and demand.

[0152] Specific behavior:

[0153] The server preprocesses the newly collected data and inputs it into the trained LSTM model, using the model.predict() method to predict the sales demand for the next month.

[0154] input:

[0155] Newly collected data, trained LSTM model.

[0156] output:

[0157] Sales trends and demand forecast results.

[0158] Step 6:

[0159] Saving and displaying prediction results

[0160] The server stores the prediction results in a database and displays them on a dashboard that users can access.

[0161] Specific behavior:

[0162] The server stores the prediction results in a database and provides them to users via a web dashboard, where users can check the prediction results.

[0163] input:

[0164] Sales trends and demand forecast results.

[0165] output:

[0166] Prediction results displayed on a dashboard.

[0167] Step 7:

[0168] Dynamic Pricing

[0169] The server dynamically sets prices based on the prediction results.

[0170] Specific behavior:

[0171] The server runs a price optimization algorithm and adjusts prices based on inventory and demand forecasts, then feeds the adjusted prices into the company's ERP system.

[0172] input:

[0173] Sales trends and demand forecast results, inventory data.

[0174] output:

[0175] Dynamically adjusted pricing information.

[0176] Step 8:

[0177] Implementing marketing strategies

[0178] The user executes an individually optimized marketing campaign based on the marketing strategy provided by the server.

[0179] Specific behavior:

[0180] The server automatically generates promotional emails and advertising campaigns and sends them to specific customer segments, and users can monitor the effectiveness of the campaigns in real time using a dashboard.

[0181] input:

[0182] Prediction results, customer data.

[0183] output:

[0184] Individually optimized marketing strategies and execution results.

[0185] These are the specific processing steps of the program for this system, which enables companies to achieve highly accurate demand forecasting, dynamic pricing, and individually optimized marketing strategies.

[0186] (Application example 1)

[0187] Next, a description will be given of Application 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."

[0188] Traditional sales trend and demand forecasting systems often lack sufficient forecasting accuracy, making them ineffective for companies' inventory management and marketing strategies. Furthermore, these systems require large desktop environments and servers and lack real-time responsiveness. As a result, it is difficult for brick-and-mortar store operators to respond quickly to changing market conditions and special seasonal events. Therefore, there is a need for a system that can achieve more accurate forecasting and dynamic pricing, thereby improving the operational efficiency of brick-and-mortar stores.

[0189] 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.

[0190] In this invention, the server includes means for collecting past sales data, external market trend data, and seasonal data, means for cleaning and preprocessing the collected data, means for training a machine learning model for demand forecasting using the preprocessed data, means for forecasting future sales trends and demand using the trained model, means for generating a dashboard for displaying the forecast results, means for dynamic pricing based on the forecast results, means for generating individually optimized marketing strategies based on customer data, and means for providing a smartphone application that performs real-time demand forecasting and dynamic pricing. This enables physical store operators to perform highly accurate demand forecasting and rapid price adaptation, thereby optimizing inventory management and marketing strategies.

[0191] "Past sales data" refers to numerical data and records relating to sales that have occurred in the past.

[0192] "External market trend data" refers to data obtained from external sources that indicates market trends and tendencies.

[0193] "Seasonal data" refers to data that indicates seasonal fluctuations and includes information such as consumer behavior at specific times.

[0194] "Means of collection" refers to the methods, tools, and processes used to obtain the necessary data.

[0195] "Cleaning" is a data preparation process that involves tasks such as deleting unnecessary elements and filling in missing values.

[0196] "Preprocessing means" refers to methods and tools used to prepare data in a format suitable for machine learning models.

[0197] "Demand forecasting" refers to estimating the future number of products that will be needed and providing information that can be incorporated into sales plans.

[0198] "Means for training a machine learning model" means the method or process of running a machine learning algorithm on collected data to build the model needed to make a prediction.

[0199] A "trained model" refers to a predictive model that has been learned by a machine learning algorithm using historical data.

[0200] "Means for generating dashboards" refers to methods for creating interfaces and tools that allow users to visually examine data.

[0201] "Dynamic pricing" is the process of adjusting prices in real time based on demand and other factors.

[0202] "Customer data" refers to information about individual customers, including purchase history and behavioral data.

[0203] An "individually optimized marketing strategy" refers to a sales promotion strategy that is customized to effectively reach a specific customer segment.

[0204] "Real-time demand forecasting" refers to systems and tools that can process data in real time and forecast demand instantly.

[0205] A "smartphone application" refers to software that runs on a mobile phone terminal and provides specific functions and services to users.

[0206] This invention provides a system that enables brick-and-mortar store operators to monitor sales trends and demand in real time and implement dynamic pricing and personalized marketing strategies. Specific embodiments for implementing this system are described below.

[0207] First, the server accesses the company's database to collect past sales data, external market trend data, and seasonal data, thereby providing the basis for highly accurate demand forecasting.

[0208] This collected data is cleaned and pre-processed on the server. Specifically, missing values ​​in the sales data are imputed using the mean and median, and outliers are removed. At this stage, the data is standardized and integrated with market trend and seasonal data.

[0209] The preprocessed data is used to train a machine learning model for demand forecasting. The server divides the data into training data and validation data, selects a time series forecasting model or neural network model, and begins training. For example, by using an LSTM (long short-term memory) model, a model that can accurately predict the next month's sales demand is built.

[0210] Furthermore, the server uses the newly collected data to forecast future sales trends and demand, and provides these forecast results to users via a dashboard, an interface that allows store operators to visually view the demand forecast results.

[0211] Users can then use these forecasts to implement dynamic pricing, which adjusts product prices in real time based on predicted demand. For example, they can maximize profits by applying discounts to items with high inventory and raising prices for items predicted to be in high demand.

[0212] Furthermore, the server automatically generates personalized marketing strategies based on the prediction results and customer data, which are tailored to specific customer segments and can, for example, send promotional emails for specific products to customers.

[0213] As part of the system, a smartphone application that performs real-time demand forecasting and dynamic pricing will also be provided, allowing store operators to easily check demand forecast results and adjust pricing on the go.

[0214] The hardware and software used are as follows: For hardware, a standard server machine and a smartphone terminal are used. For software, Python, Pandas, NumPy, Scikit-learn, Keras, etc. are used. Using these tools, data collection, cleaning, preprocessing, model training, and prediction can be performed efficiently.

[0215] A concrete example is a scenario in which the system needs to predict the demand for each product this Christmas season based on sales figures from last December, and then make pricing suggestions based on that. The user can request this by entering the following prompt:

[0216] "Based on last December's sales figures, please predict the demand for each product during this year's Christmas season. Also, please make pricing suggestions based on your results."

[0217] In this way, the present invention not only dramatically improves the operational efficiency of physical stores, but also realizes highly accurate demand forecasting and dynamic price adjustments.

[0218] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0219] Step 1:

[0220] A user launches a smartphone application and submits a request to collect historical sales data, external market trend data, and seasonality data.

[0221] Input: Request data

[0222] Output: Collection request

[0223] Specific Operation: When a user presses a button on the application to start data collection, a request is sent to the server to collect historical sales data, external market trend data, and seasonal data.

[0224] Step 2:

[0225] The server accesses the company's database to retrieve historical sales data, collects market trend data from external APIs, and loads pre-defined seasonality data.

[0226] Input: Collection request

[0227] Output: Raw data

[0228] What happens: The server queries the company's database to retrieve historical sales data, and simultaneously sends a request to an external API to retrieve market trend data and load seasonal data files.

[0229] Step 3:

[0230] The server cleans and pre-processes the collected data, imputing missing values, removing outliers, and standardizing the data. It also integrates market trend and seasonal data with sales data.

[0231] Input: Raw data

[0232] Output: Preprocessed data

[0233] What it does: It uses Pandas to impute missing data values ​​with the mean or median, detect and remove outliers, and standardize and combine data into a single dataset.

[0234] Step 4:

[0235] The server uses the preprocessed data to train a machine learning model for demand forecasting. It splits the data into training data and validation data and uses a time series model such as LSTM.

[0236] Input: Preprocessed data

[0237] Output: A trained model

[0238] Specific operation: Using Scikit-learn and Keras, the data is divided into training data and validation data, and an LSTM model is constructed and trained.

[0239] Step 5:

[0240] The server uses the newly collected data to predict future sales trends and demand with a trained model.

[0241] Input: New raw data

[0242] Output: Demand forecast results

[0243] Specific operation: After preprocessing the newly collected sales data and market trend data, it is input into the trained model to predict future demand.

[0244] Step 6:

[0245] The server dynamically sets prices using a pricing algorithm based on the prediction results.

[0246] Input: Demand forecast results

[0247] Output: Pricing data

[0248] Specific operation: Based on the prediction results, the server performs dynamic pricing, lowering the prices of items with high inventory and raising the prices of items with high demand.

[0249] Step 7:

[0250] The forecast results and pricing data are generated as a dashboard and displayed on a smartphone application.

[0251] Input: Forecast results, pricing data

[0252] Output: Dashboard

[0253] Specific operation: The server visualizes the forecast results and pricing data, generates a dashboard, and delivers it to a smartphone application.

[0254] Step 8:

[0255] Users can check the dashboard through a smartphone application and adjust prices and manage inventory in physical stores based on sales strategies.

[0256] Input: Dashboard

[0257] Output: Sales strategy implementation

[0258] Specific Action: User reviews the application dashboard and implements specific sales strategies based on predicted demand and pricing.

[0259] Through this process, physical store operators can achieve highly accurate demand forecasts and quick price adjustments.

[0260] 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.

[0261] As an embodiment of the present invention, we provide a system that uses AI and machine learning to predict future sales trends and demand, and generates marketing strategies based on user emotions using an emotion engine. This system utilizes past sales data, market trend data, seasonal data, and customer emotion data to enable companies to effectively optimize sales plans, inventory management, and marketing strategies.

[0262] First, the server accesses the company's database to collect past sales data, external market trend data, and seasonal data. For example, it obtains detailed information such as monthly sales data for the past five years and the sales volume and price of each product. It also obtains market trend data from an external API. It also accesses a calendar event database to collect data on seasonal events such as Christmas and Valentine's Day.

[0263] The server then cleans and preprocesses the collected data. For example, it imputes missing values ​​in the sales data using the mean or median, and removes outliers. This preprocessing phase also scales the data and integrates sales data, market trend data, and seasonality data. The resulting dataset is then used to train machine learning models.

[0264] The server then uses the preprocessed data to train a machine learning model. Here, the data is split into training data and validation data, and a time series forecasting model or neural network model is selected and trained. For example, the server uses an LSTM (long short-term memory) model to train a model to predict sales demand for the next month. During the training process, the data is repeatedly fed into the model, and patterns are learned.

[0265] The system also incorporates an emotion engine that analyzes user emotions and optimizes marketing strategies based on those emotions. Specifically, the server uses customer data to analyze customer emotions in real time. For example, if a customer expresses positive emotions, the system may introduce new products or make cross-selling suggestions. Conversely, if a customer expresses negative emotions, the system may provide customer support or send discount coupons.

[0266] Once training is complete, the server uses the newly collected data to predict future sales trends and demand. These predictions are stored locally and then made available to users via a dashboard. For example, the dashboard might display sales forecasts for a specific product during the upcoming Christmas season.

[0267] Once the forecast results are available, the server then performs dynamic pricing, adjusting product prices in real time based on predicted demand. For example, discounts are applied to items with high inventory, and prices are increased for items predicted to be in high demand. This dynamic pricing allows businesses to maximize profits.

[0268] Furthermore, the device displays the forecast results to the user through a dashboard. Information is presented using visual elements (graphs, charts) for easy understanding. Users can also monitor the effectiveness of their marketing efforts in real time, allowing them to quickly determine the success of their efforts and any necessary adjustments.

[0269] As described above, the present invention provides a system that provides dynamic pricing based on highly accurate demand forecasts and an optimized marketing strategy using an emotion engine, contributing to corporate sales planning, inventory management, and profit maximization.

[0270] The processing flow will be explained below.

[0271] Step 1:

[0272] The server accesses the company's database and collects past sales data, such as monthly sales data for the past five years and detailed information such as the sales volume and price of each product.

[0273] Step 2:

[0274] The server uses external APIs to collect market trend data, specifically, Google Trends and social media data, to understand consumer interest trends in the market.

[0275] Step 3:

[0276] The server accesses a calendar event database and collects data that reflects seasonality, for example, collecting data about seasonal events such as Christmas and Valentine's Day.

[0277] Step 4:

[0278] The server cleans the collected data by detecting missing values ​​in the dataset, imputing them with the mean or median, and removing outliers.

[0279] Step 5:

[0280] The server pre-processes the cleaned data, including scaling and normalizing it, and integrating sales data, market trend data, and seasonality data.

[0281] Step 6:

[0282] The server splits the preprocessed dataset into training and validation data, typically 80% of the total data for training and 20% for validation.

[0283] Step 7:

[0284] The server selects a time series prediction model or neural network model, for example, a long short-term memory (LSTM) model.

[0285] Step 8:

[0286] The server uses the training data to train the machine learning model by repeatedly inputting the data into the model and letting it learn patterns.

[0287] Step 9:

[0288] The server evaluates the performance of the trained model using validation data, using metrics such as MSE (Mean Squared Error) and MAE (Mean Absolute Error).

[0289] Step 10:

[0290] The server uses the new data to predict future sales trends and demand, for example, predicting the sales volume of a particular product for the next month.

[0291] Step 11:

[0292] The server stores the prediction results in a database and generates a dashboard, which displays the prediction results in the form of graphs and charts.

[0293] Step 12:

[0294] The terminal displays the prediction results to the user through a dashboard, providing visual information that is easy for the user to understand.

[0295] Step 13:

[0296] The server dynamically sets prices based on the demand forecast results, for example, raising prices for products with high predicted demand and discounting products with high inventory.

[0297] Step 14:

[0298] The user implements the individually optimized marketing strategy provided by the server, specifically, sending personalized promotional emails to specific customer segments.

[0299] Step 15:

[0300] The server runs an emotion engine using customer data, such as past purchase history, browsing history, and social media comments, to analyze customer emotions in real time.

[0301] Step 16:

[0302] The server uses the emotion engine to generate more effective marketing strategies based on the obtained emotion data, such as introducing new products to customers who show positive emotions and providing support or discount coupons to customers who show negative emotions.

[0303] Step 17:

[0304] Users can monitor the effectiveness of their marketing initiatives in real time, allowing them to quickly determine the success of their initiatives and make any necessary adjustments.

[0305] Example 2

[0306] Next, a description will be given of 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."

[0307] For businesses, accurate demand forecasting and determining effective marketing strategies are important issues that directly affect sales growth and inventory management. However, many businesses are unable to properly utilize past data, making it difficult to accurately reflect seasonality and market trends. Formulating appropriate marketing strategies based on customer sentiment is also a challenge. The present invention aims to solve these problems and provide a set of tools that enable businesses to optimize sales planning, inventory management, and marketing strategies.

[0308] 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.

[0309] In this invention, the server includes means for collecting past transaction data, external market trend data, and seasonal information, means for cleaning and pre-processing the collected data, means for training a machine learning model for demand forecasting using the pre-processed data, means for forecasting future sales trends and demand using the trained model, means for generating a visualization interface for displaying the forecast results, means for dynamic pricing based on the forecast results, and means for analyzing customer sentiment data using a sentiment analyzer and generating individually optimized marketing strategies based on the results, thereby enabling highly accurate sales forecasting and optimization of marketing strategies based on customer sentiment.

[0310] "Past transaction data" refers to information about all sales activities conducted by a company in the past, including details such as sales quantity, sales price, and sales date and time.

[0311] "External market trend data" refers to information that shows overall market trends, competitor activities, consumer purchasing trends, etc., and is obtained from external information providers or APIs.

[0312] "Seasonal information" is data that indicates fluctuations in sales related to specific seasons or events, and includes sales trends during specific periods such as Christmas and Valentine's Day.

[0313] "Cleaning" refers to the process of removing errors, missing values, and inappropriate data from collected data to improve the quality of the data.

[0314] "Preprocessing" is the process of preparing data in a format suitable for a machine learning model, and includes scaling and standardizing the data, and filling in missing values.

[0315] A "machine learning model" is an algorithm or system used to make predictions, classifications, and optimizations based on input data, and includes LSTM and neural networks.

[0316] "Training" refers to the process of using existing data to teach a machine learning model so that it can make appropriate predictions and recommendations for new data.

[0317] "Future Sales Trends" refers to future sales trends and changes in demand predicted by the machine learning model.

[0318] "Prediction results" refer to the results of a machine learning model's predictions about the future, and include information such as specific sales volumes and demand patterns.

[0319] A "visualization interface" is a means for displaying prediction results in a visual format such as a graph or chart, making it easy for users to understand.

[0320] "Dynamic pricing" refers to the practice of adjusting product prices in real time based on predicted demand, raising or lowering prices based on inventory levels and supply-demand balances.

[0321] An "emotion analyzer" is a device or software that analyzes customer emotions and determines the emotional state of customers by analyzing reviews, social media posts, etc.

[0322] An "individually optimized marketing strategy" refers to a strategy that designs and executes optimal marketing methods for individual customers based on collected customer data and the results of sentiment analysis.

[0323] This invention is a system that uses AI and machine learning to predict future sales trends and demand, and further uses an emotion engine to generate marketing strategies based on user emotions. This system aims to help companies optimize their sales plans, inventory management, and marketing strategies.

[0324] Data collection

[0325] server

[0326] The server first accesses the company's database to retrieve transaction data from the past five years. This data includes details such as the sales quantity, sales price, and sales date for each product. It also collects market trend data through an external API. This market trend data includes consumer purchasing trends and competitor activity. The server also accesses a calendar event database to collect information about seasonal events such as Christmas and Valentine's Day.

[0327] Data Preprocessing

[0328] server

[0329] The server cleans and preprocesses the collected data. Specifically, if there are missing values ​​in the sales data, they are imputed using the mean or median. Outliers are also removed as soon as they are detected. The data is then scaled and converted into a unified format. Sales data, market trend data, and seasonal data are integrated into a single dataset.

[0330] Training a machine learning model

[0331] server

[0332] The preprocessed dataset is used to train the machine learning model. The dataset is divided into training data and validation data. The server selects an LSTM (long short-term memory) model and trains it to learn sales demand by inputting data into the model. This process is performed so that the model can learn sales patterns and improve its prediction accuracy for future data.

[0333] Predicting future sales trends

[0334] server

[0335] Once training is complete, the newly collected data is used to predict future sales trends and demand, such as predicting sales volumes for a particular product during the upcoming Christmas season. These predictions are stored on the server and later provided to users via a dashboard.

[0336] Utilizing the Emotion Engine

[0337] server

[0338] The server uses a sentiment analyzer to analyze user sentiment in real time. For example, it analyzes customer reviews and social media posts, and if the sentiment is positive, it introduces new products or suggests cross-selling. Conversely, if the sentiment is negative, it provides customer support or sends discount coupons.

[0339] Dynamic Pricing

[0340] server

[0341] Based on the predictions, the server dynamically sets prices, for example, applying discounts to items with high inventory and raising prices for items predicted to be in high demand, aiming to maximize the company's profits.

[0342] Viewing the Dashboard

[0343] Terminal

[0344] The dashboard displays the forecast results sent from the server, using visual elements such as graphs and charts to provide information that users can easily understand. Users can also monitor the effectiveness of their marketing efforts in real time, quickly determining the success of their efforts and any necessary adjustments.

[0345] Specific examples

[0346] For example, the server collects data for the Christmas shopping season in December. First, it collects sales data from past Christmas seasons from the database, then obtains market trend data using an external API. It also collects Christmas-related data from the calendar event database. The server then imputes any missing values ​​in the collected data using the average value, scales and integrates all the data. Once all the datasets are ready, it uses them to train an LSTM model. It then performs predictions on the newly acquired data and displays the results on a dashboard. Finally, it uses an emotion engine to introduce new products if customers have positive emotions.

[0347] Prompt Sentence Examples

[0348] "For the Christmas sales season leading up to December, collect Christmas season sales data, market trend data, and seasonal event data from the past five years, and use this data to make predictions using an LSTM model and display them on a dashboard. In addition, analyze customer sentiment data in real time, and if customers show positive sentiment, introduce new products or make cross-selling suggestions."

[0349] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0350] Step 1: Data collection

[0351] server

[0352] The server first accesses the company's database and retrieves transaction data from the past five years, including details such as the quantity, price, and date of each product sold.

[0353] Input: Company database

[0354] Output: Transaction data for the past 5 years

[0355] The server then collects market trend data through external APIs, including consumer purchasing habits and competitor activity.

[0356] Input: External API

[0357] Output: Market trend data

[0358] Additionally, the server accesses a calendar events database to collect information about seasonal events such as Christmas and Valentine's Day.

[0359] Input: Calendar event database

[0360] Output: Seasonal event data

[0361] Step 2: Data Preprocessing

[0362] server

[0363] The server cleans and preprocesses all collected data. Specifically, if there are missing values ​​in the sales data, they are imputed using the mean or median. Outliers are also removed as soon as they are detected. The server also scales the data and converts it into a unified format. Finally, the sales data, market price trend data, and seasonal data are integrated into a single dataset.

[0364] Inputs: trade data, market trend data, seasonal event data

[0365] Output: Preprocessed dataset

[0366] Step 3: Train the model

[0367] server

[0368] The preprocessed dataset is used to train a machine learning model. The dataset is divided into training data and validation data. The server selects an LSTM (long short-term memory) model and inputs the data multiple times to train the model on sales demand. This training process allows the model to learn sales patterns and improve its prediction accuracy for future data.

[0369] Input: Preprocessed dataset

[0370] Output: A trained LSTM model

[0371] Step 4: Predict future sales trends

[0372] server

[0373] Using the trained LSTM model, the server analyzes newly collected data and predicts future sales trends and demand. For example, it predicts the sales volume of a particular product during the upcoming Christmas season. The prediction results are stored on the server and later provided to users via a dashboard.

[0374] Input: Newly collected data, trained LSTM model

[0375] Output: Future sales trend forecast results

[0376] Step 5: Leverage the Emotion Engine

[0377] server

[0378] The server uses a sentiment analyzer to analyze user emotions in real time. Specifically, it analyzes customer reviews and social media posts, and introduces new products and makes cross-selling suggestions to customers who show positive emotions. Conversely, it provides customer support and sends discount coupons to customers who show negative emotions.

[0379] Input: Customer reviews, social media posts

[0380] Output: Emotion-based marketing strategies

[0381] Step 6: Dynamic Pricing

[0382] server

[0383] Based on the predictions, the server dynamically sets prices, applying discounts to items with high inventory and raising prices for items predicted to be in high demand, thereby maximizing profits for the company.

[0384] Input: Future sales trend forecast results

[0385] Output: Real-time adjusted product prices

[0386] Step 7: View the dashboard

[0387] Terminal

[0388] The dashboard displays the prediction results sent from the server. The dashboard uses visual elements (graphs, charts) to provide users with easy-to-understand information. Users can also monitor the effectiveness of their marketing initiatives in real time through the dashboard, allowing them to quickly determine the success of their initiatives and any necessary adjustments.

[0389] Input: Future sales trend prediction results, emotion-based marketing strategies

[0390] Output: Visually displayed forecast results and strategies

[0391] (Application example 2)

[0392] Next, a description will be given of Application 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."

[0393] Modern marketing strategies and sales plans must respond quickly and accurately to diverse customer needs. However, traditional systems are based solely on past sales data and market trend data, which means they cannot fully consider customer sentiment or real-time conditions. Other issues include the difficulty of improving the accuracy of demand forecasts and dynamic pricing, and providing individually optimized marketing strategies in real time. This makes it difficult for companies to optimize sales plans and inventory management, leading to a demand for systems that contribute to maximizing profits.

[0394] 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.

[0395] In this invention, the server includes means for collecting past sales data, external market trend data, and seasonal data, means for cleaning and pre-processing the collected data, means for training a machine learning model for demand forecasting using the pre-processed data, means for collecting and analyzing customer sentiment data, and means for optimizing marketing strategies based on customer sentiment, thereby improving the accuracy of demand forecasting and marketing strategies and enabling optimal sales planning and inventory management that take customer sentiment into account.

[0396] "Sales data" refers to historical data relating to past sales, inventory information, product sales quantities and prices, and the like.

[0397] "Market trend data" refers to data on customer purchasing trends, popular product trends, and economic conditions in a broad range of markets.

[0398] "Seasonal data" refers to data on fluctuations in customer purchasing behavior associated with specific seasons or events, such as Christmas or Valentine's Day.

[0399] A "machine learning model" is a collection of algorithms that learn patterns from data and make predictions or classifications based on those patterns.

[0400] "Emotional data" refers to data on the emotional movements and feelings of customers that can be obtained from their facial expressions, voice, and behavior.

[0401] A "marketing strategy" is a plan for a series of advertising, promotion, pricing, and other activities designed to achieve a specific objective.

[0402] "Training data" is a learning dataset used when building a machine learning model.

[0403] "Validation data" is a dataset used to evaluate the performance of a trained machine learning model.

[0404] A "dashboard" is an interface that visually displays forecast results and other important information in a format that is easy for users to understand.

[0405] "Dynamic pricing" is a method of adjusting product prices in real time based on demand forecasts and inventory status.

[0406] "Demand forecasting" refers to predicting future customer purchasing behavior and sales volumes based on past data and trend information.

[0407] As an embodiment of the present invention, a server uses AI and machine learning to predict future sales trends and demand, and also uses an emotion engine to generate a marketing strategy based on user emotions. Specific implementation means are described below.

[0408] Data collection

[0409] The server accesses the company's database to collect past sales data, external market trend data, and seasonal data, including the "sales data," "market trend data," and "seasonal data" previously specified. For example, the server uses an API to obtain external trend data and retrieves past sales history from the company's internal database.

[0410] Data Cleaning and Preprocessing

[0411] The collected data is cleaned, missing values ​​are filled, and outliers are removed. This is a necessary step to increase the reliability of the data. For example, missing values ​​are filled with the mean or median. The data is then scaled and sales data, market trend data, and seasonal data are integrated.

[0412] Training a machine learning model

[0413] The preprocessed data is used to train a machine learning model. The model used here is a "long short-term memory (LSTM) model" or a "neural network model." The server splits the data into training data and validation data and performs training.

[0414] Demand forecasting

[0415] The trained model is used to predict future sales trends and demand, for example, predicting demand for specific products around seasonal events like Christmas and Valentine's Day. These predictions are then displayed on a dashboard as graphs and charts that users can easily understand.

[0416] Emotion data collection and analysis

[0417] The server collects customer emotional data from the smartphone's camera and microphone. Using specialized software for emotion analysis, it analyzes facial expressions and voice in real time. This emotional data is used to optimize marketing strategies.

[0418] Optimizing your marketing strategy

[0419] Optimize your marketing strategies based on customer sentiment data. For example, if a customer expresses positive sentiment, introduce new products or offer cross-selling suggestions. If a customer expresses negative sentiment, offer customer support or send discount coupons.

[0420] Dynamic Pricing

[0421] Dynamically adjust pricing based on demand forecasts. If demand for a particular product is predicted to increase, prices are raised, while discounts are offered if inventory is high. This allows for real-time price adjustments to maximize profits.

[0422] Dashboard View

[0423] It generates a dashboard that visually displays forecast results, sentiment analysis results, and pricing information, allowing users to easily check upcoming sales plans and marketing initiatives and quickly adjust them as needed.

[0424] For example, the following prompt sentence is input to the generative AI model:

[0425] "Please forecast the demand for home video game consoles during the next Christmas season."

[0426] "Analyze recent user reviews to identify products that have received positive feedback."

[0427] This allows companies to develop highly accurate sales strategies based on market trends and customer sentiment, and implement optimized marketing strategies.

[0428] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0429] Step 1:

[0430] The server collects historical sales data from the company's database and obtains external market trend data and seasonal data through API. The input is database connection information and API key, and the output is the integrated raw data set. Specifically, it issues database queries and API requests and stores the obtained data in local storage.

[0431] Step 2:

[0432] The server performs data cleaning on the collected raw dataset. The input is the integrated raw dataset, and the output is a clean dataset. Specific operations include imputing missing values ​​(calculating the mean and median), removing outliers, and performing data scaling. This results in a dataset suitable for training machine learning models.

[0433] Step 3:

[0434] The server trains a machine learning model using a clean dataset. The input is the clean dataset, and the output is the trained model. Specifically, the data is split into training data and validation data, and training is performed using an LSTM model or a neural network model. Machine learning libraries such as TensorFlow and Keras are used to train the model.

[0435] Step 4:

[0436] The server uses the trained model to predict future sales trends and demand. The input is a newly collected dataset, and the output is the prediction result. Specifically, new data is input into the model, and predictions are made to calculate demand fluctuations and sales trends.

[0437] Step 5:

[0438] The device collects user emotional data through the smartphone's camera and microphone. The input is video and audio data, and the output is the result of emotional analysis. Specifically, it uses an emotion analysis engine to analyze changes in facial expressions and voice tone in real time to extract emotional data.

[0439] Step 6:

[0440] The server generates an individually optimized marketing strategy based on the emotional data. The input is the result of the emotional analysis, and the output is a set of marketing measures. Specifically, it runs an algorithm that performs promotions and product recommendations according to the emotional state.

[0441] Step 7:

[0442] The server performs dynamic pricing based on the demand forecast results. The inputs are the demand forecast results and inventory data, and the output is adjusted price information. Specifically, it runs an algorithm that raises prices when demand is high and applies discounts when inventory is high.

[0443] Step 8:

[0444] The terminal generates a dashboard and displays the forecast results, sentiment analysis results, and pricing information. The input is the forecast results and sentiment analysis results, and the output is a visually displayed dashboard. Specific operations include displaying information using graphs and charts, allowing users to easily check sales forecasts and marketing strategies.

[0445] Through these steps, the server and terminals work together to provide a system that realizes highly accurate demand forecasting, dynamic pricing, and emotion-based marketing strategies.

[0446] 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.

[0447] 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> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) 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.

[0448] 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.

[0449] [Second embodiment]

[0450] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0451] 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.

[0452] 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).

[0453] 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.

[0454] 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.

[0455] 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).

[0456] 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.

[0457] 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.

[0458] 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.

[0459] 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.

[0460] 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.

[0461] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0462] As an embodiment of the present invention, we provide a system that uses AI and machine learning to forecast future sales trends and demand. This system aims to enable companies to effectively optimize sales planning, inventory management, and marketing strategies by using past sales data, market trend data, and seasonal data.

[0463] The specific program processing of this system is as follows.

[0464] First, the server accesses the company's database to collect past sales data, external market trend data, and seasonal data. For example, the server obtains monthly sales data for the past five years and then obtains market trend data from an external API. In this way, multifaceted data collection forms the basis for a more accurate forecasting model.

[0465] The server then cleans and preprocesses the collected data. For example, missing values ​​in the sales data are imputed using the mean or median. Outliers are also removed. This preprocessing phase normalizes the data and integrates market trend and seasonal data with the sales data. The resulting dataset is then used to train the machine learning model.

[0466] The server then uses the preprocessed data to train a machine learning model. Here, the data is split into training data and validation data, and a time series forecasting model or neural network model is selected and trained. For example, the server uses an LSTM (long short-term memory) model to train a model to predict sales demand for the next month. During this training process, the data is repeatedly fed into the model, and patterns are learned.

[0467] Once training is complete, the server uses the newly collected data to predict future sales trends and demand. These predictions are stored locally and then made available to users via a dashboard. For example, the dashboard might display sales forecasts for a specific product during the upcoming Christmas season.

[0468] Once the forecast results are available, the server then performs dynamic pricing, adjusting product prices in real time based on predicted demand. For example, discounts are applied to items with high inventory, and prices are increased for items predicted to be in high demand. This dynamic pricing allows businesses to maximize profits.

[0469] Furthermore, the user implements individually optimized marketing strategies provided by the server. The server generates personalized promotional emails for specific customer segments based on the prediction results and customer data. The user can then run advertising campaigns based on this and monitor the results in real time.

[0470] In this way, the present invention provides dynamic pricing and optimized marketing strategies based on highly accurate demand forecasts, realizing a system that contributes to corporate sales planning, inventory management, and profit maximization.

[0471] The processing flow will be explained below.

[0472] Step 1:

[0473] The server accesses the company's database and collects past sales data, such as monthly sales data for the past five years and detailed information such as the sales volume and price of each product.

[0474] Step 2:

[0475] The server uses external APIs to collect relevant market trend data, specifically Google Trends and social media data, to understand consumer interest trends in the market.

[0476] Step 3:

[0477] The server accesses a calendar event database to retrieve data that reflects seasonality, for example, collecting data about seasonal events such as Christmas and Valentine's Day.

[0478] Step 4:

[0479] The server cleans the collected data by detecting missing values ​​in the dataset, imputing them with the mean or median, and removing outliers.

[0480] Step 5:

[0481] The server pre-processes the cleaned data, including scaling and normalizing the data, and integrating sales data, market trend data, and seasonality data.

[0482] Step 6:

[0483] The server splits the preprocessed dataset into training and validation data, typically 80% of the total data for training and 20% for validation.

[0484] Step 7:

[0485] The server selects a time series prediction model or neural network model, for example, a long short-term memory (LSTM) model.

[0486] Step 8:

[0487] The server uses the training data to train the machine learning model, which involves repeatedly feeding the data into the model and letting it learn patterns.

[0488] Step 9:

[0489] The server evaluates the performance of the trained model using validation data, using metrics such as MSE (Mean Squared Error) and MAE (Mean Absolute Error).

[0490] Step 10:

[0491] The server uses the new data to predict future sales trends and demand, for example, predicting the sales volume of a particular product for the next month.

[0492] Step 11:

[0493] The server stores the prediction results in a database and generates a dashboard, which displays the prediction results in the form of graphs and charts.

[0494] Step 12:

[0495] The device displays the prediction results to the user through a dashboard, providing visual information that is easy for the user to understand.

[0496] Step 13:

[0497] The server dynamically sets prices based on the demand forecast results, for example, raising prices for products with high predicted demand and discounting products with high inventory.

[0498] Step 14:

[0499] The user implements individually optimized marketing strategies generated by the server, for example, sending personalized promotional emails to specific customer segments.

[0500] Step 15:

[0501] Users can monitor the effectiveness of their marketing efforts in real time, allowing them to quickly determine the success of their efforts and make any necessary adjustments.

[0502] Example 1

[0503] Next, a description will be given of 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."

[0504] Conventional sales forecasting systems often base their forecasts solely on past sales data and are unable to comprehensively incorporate external market trend data or seasonal data, resulting in low forecast accuracy. Furthermore, they lack effective means for implementing dynamic pricing and individually optimized marketing strategies. A system that can solve this issue and provide highly accurate demand forecasts and dynamic pricing and marketing strategies based on those forecasts is needed.

[0505] 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.

[0506] In this invention, the server includes means for collecting past sales data, external market trend data, and seasonal data, means for cleaning and pre-processing the collected data, means for training a machine learning model for demand forecasting using the pre-processed data, means for forecasting future sales trends and demand using the trained model, means for generating a recording medium for displaying the forecast results, means for dynamic pricing based on the forecast results, and means for generating an individually optimized marketing strategy based on customer data, thereby enabling highly accurate demand forecasting, dynamic pricing, and the implementation of individually optimized marketing strategies.

[0507] "Sales data" is a collection of information about sales performance of a particular product over a past period.

[0508] "Market trend data" is a collection of information that indicates the current situation and future trends in a particular market or industry.

[0509] "Seasonal data" is a collection of information that indicates sales trends and demand fluctuations associated with particular seasons or annual events.

[0510] "Data cleaning" is the process of removing inaccurate, incomplete, or unnecessary values ​​from collected data to improve data quality.

[0511] "Preprocessing" refers to a series of steps taken to convert collected data into a format suitable for analysis and machine learning.

[0512] A "machine learning model" is an algorithm or mathematical model that learns patterns from data and performs tasks such as prediction and classification.

[0513] "Training" is the process of providing a machine learning model with large amounts of data and having it learn from that data to improve its prediction accuracy.

[0514] A "recording medium" is a physical or electronic device or means for storing and displaying digital data.

[0515] "Dynamic pricing" is a method of adjusting prices in real time based on demand forecasts and market conditions.

[0516] A "marketing strategy" is a series of plans and measures implemented to promote sales and increase brand awareness among specific customer segments.

[0517] This invention relates to a sales forecasting system, specifically a system that uses AI and machine learning technologies to forecast future sales trends and demand. This system is designed to help companies more effectively plan and execute sales plans, inventory management, and marketing strategies by utilizing past sales data, market trend data, and seasonal data.

[0518] Hardware and software used

[0519] 1. Server: Manages all data collection, preprocessing, machine learning model training, and dynamic pricing using libraries such as Python, Pandas, Scikit-learn, TensorFlow, or PyTorch.

[0520] 2. Database: Stores and manages company sales data and external trend data. Specifically, MySQL or PostgreSQL can be used.

[0521] 3. External API: Used to obtain market trend data.

[0522] 4. Web dashboard: An interface for providing predictions and analysis results to users. Specifically, JavaScript, D3.js, or Tableau may be used.

[0523] Data collection and preprocessing

[0524] The server accesses the company's database to collect past sales data, external market trend data, and seasonal data. This data allows for more accurate forecasts based on multifaceted information. Specifically, the server periodically sends API requests to collect and store the necessary data.

[0525] Next, the server cleans and preprocesses the data. This step includes imputing missing values, removing outliers, and normalizing the data. For example, it uses the Pandas fillna() method to impute missing values ​​with the mean.

[0526] Training a machine learning model

[0527] The server uses the preprocessed data to train a machine learning model. Here, the data is divided into training data and validation data, and a long short-term memory (LSTM) model is used to perform time series prediction. Specifically, an LSTM model is built and trained using TensorFlow.

[0528] Future sales trends and demand forecasts

[0529] Using the trained model, the server predicts future sales trends and demand based on newly collected data. These predictions are stored in a database and made available to users through a web dashboard. For example, the dashboard might predict that demand for a particular product will increase during the upcoming Christmas season.

[0530] Dynamic Pricing

[0531] The server then dynamically sets prices based on the forecast, for example, discounting items with high inventory and raising prices for items predicted to be in high demand. This dynamic pricing is a key factor in helping companies maximize profits.

[0532] Individually optimized marketing strategies

[0533] Based on the prediction results and customer data provided by the server, users can create personalized promotional emails and advertising campaigns for specific customer segments. For example, emails can be automatically generated to strongly appeal to specific customer demographics about a specific product.

[0534] Prompt Sentence Examples

[0535] "Please provide us with your sales forecast for electronic products for the upcoming Christmas season. Also, please identify which products will sell best and suggest a corresponding marketing strategy."

[0536] In this way, this system provides dynamic pricing and optimized marketing strategies based on highly accurate demand forecasts, contributing to corporate sales planning, inventory management, and profit maximization.

[0537] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0538] Step 1:

[0539] Data collection

[0540] The server accesses the company's database to collect historical sales data, external market trend data, and seasonality data.

[0541] Specific behavior:

[0542] The server periodically executes SQL queries to extract sales data from the company's database. The server sends HTTP requests to retrieve market trend data from an external API. The server stores this data in a local database.

[0543] input:

[0544] Corporate database connection information, external API endpoints.

[0545] output:

[0546] Collection of historical sales data, market trend data, and seasonality data.

[0547] Step 2:

[0548] Data Cleaning and Preprocessing

[0549] The server cleans the collected data, filling in missing values, removing outliers, and normalizing it.

[0550] Specific behavior:

[0551] The server uses the Pandas library to load the data into a data frame, uses the fillna() method to impute missing values, boxplots and Z-scores to detect and remove outliers, and normalizes each data set to create a unified dataset.

[0552] input:

[0553] Collected sales data, market trend data, and seasonality data.

[0554] output:

[0555] Cleaned and preprocessed integrated dataset.

[0556] Step 3:

[0557] Data partitioning

[0558] The server splits the combined dataset into training data and validation data.

[0559] Specific behavior:

[0560] The server splits the data into training and validation data using Scikit-learn's train_test_split function.

[0561] input:

[0562] Cleaned and preprocessed integrated dataset.

[0563] output:

[0564] Training and validation data.

[0565] Step 4:

[0566] Training a machine learning model

[0567] The server trains the LSTM model using the training data.

[0568] Specific behavior:

[0569] The server uses the TensorFlow library to build an LSTM model and trains it using the model.fit() method. During the training process, the validation data is used to evaluate the accuracy of the model.

[0570] input:

[0571] Training data, validation data.

[0572] output:

[0573] A trained LSTM model.

[0574] Step 5:

[0575] Future predictions

[0576] The server uses the trained model to predict future sales trends and demand.

[0577] Specific behavior:

[0578] The server preprocesses the newly collected data and inputs it into the trained LSTM model, using the model.predict() method to predict the sales demand for the next month.

[0579] input:

[0580] Newly collected data, trained LSTM model.

[0581] output:

[0582] Sales trends and demand forecast results.

[0583] Step 6:

[0584] Saving and displaying prediction results

[0585] The server stores the prediction results in a database and displays them on a dashboard that users can access.

[0586] Specific behavior:

[0587] The server stores the prediction results in a database and provides them to users via a web dashboard, where users can check the prediction results.

[0588] input:

[0589] Sales trends and demand forecast results.

[0590] output:

[0591] Prediction results displayed on a dashboard.

[0592] Step 7:

[0593] Dynamic Pricing

[0594] The server dynamically sets prices based on the prediction results.

[0595] Specific behavior:

[0596] The server runs a price optimization algorithm and adjusts prices based on inventory and demand forecasts, then feeds the adjusted prices into the company's ERP system.

[0597] input:

[0598] Sales trends and demand forecast results, inventory data.

[0599] output:

[0600] Dynamically adjusted pricing information.

[0601] Step 8:

[0602] Implementing marketing strategies

[0603] The user executes an individually optimized marketing campaign based on the marketing strategy provided by the server.

[0604] Specific behavior:

[0605] The server automatically generates promotional emails and advertising campaigns and sends them to specific customer segments, and users can monitor the effectiveness of the campaigns in real time using a dashboard.

[0606] input:

[0607] Prediction results, customer data.

[0608] output:

[0609] Individually optimized marketing strategies and execution results.

[0610] These are the specific processing steps of the program for this system, which enables companies to achieve highly accurate demand forecasting, dynamic pricing, and individually optimized marketing strategies.

[0611] (Application example 1)

[0612] Next, a description will be given of Application 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."

[0613] Traditional sales trend and demand forecasting systems often lack sufficient forecasting accuracy, making them ineffective for companies' inventory management and marketing strategies. Furthermore, these systems require large desktop environments and servers and lack real-time responsiveness. As a result, it is difficult for brick-and-mortar store operators to respond quickly to changing market conditions and special seasonal events. Therefore, there is a need for a system that can achieve more accurate forecasting and dynamic pricing, thereby improving the operational efficiency of brick-and-mortar stores.

[0614] 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.

[0615] In this invention, the server includes means for collecting past sales data, external market trend data, and seasonal data, means for cleaning and preprocessing the collected data, means for training a machine learning model for demand forecasting using the preprocessed data, means for forecasting future sales trends and demand using the trained model, means for generating a dashboard for displaying the forecast results, means for dynamic pricing based on the forecast results, means for generating individually optimized marketing strategies based on customer data, and means for providing a smartphone application that performs real-time demand forecasting and dynamic pricing. This enables physical store operators to perform highly accurate demand forecasting and rapid price adaptation, thereby optimizing inventory management and marketing strategies.

[0616] "Past sales data" refers to numerical data and records relating to sales that have occurred in the past.

[0617] "External market trend data" refers to data obtained from external sources that indicates market trends and tendencies.

[0618] "Seasonal data" refers to data that indicates seasonal fluctuations and includes information such as consumer behavior at specific times.

[0619] "Means of collection" refers to the methods, tools, and processes used to obtain the necessary data.

[0620] "Cleaning" is a data preparation process that involves tasks such as deleting unnecessary elements and filling in missing values.

[0621] "Preprocessing means" refers to methods and tools used to prepare data in a format suitable for machine learning models.

[0622] "Demand forecasting" refers to estimating the future number of products that will be needed and providing information that can be incorporated into sales plans.

[0623] "Means for training a machine learning model" means the method or process of running a machine learning algorithm on collected data to build the model needed to make a prediction.

[0624] A "trained model" refers to a predictive model that has been learned by a machine learning algorithm using historical data.

[0625] "Means for generating dashboards" refers to methods for creating interfaces and tools that allow users to visually examine data.

[0626] "Dynamic pricing" is the process of adjusting prices in real time based on demand and other factors.

[0627] "Customer data" refers to information about individual customers, including purchase history and behavioral data.

[0628] An "individually optimized marketing strategy" refers to a sales promotion strategy that is customized to effectively reach a specific customer segment.

[0629] "Real-time demand forecasting" refers to systems and tools that can process data in real time and forecast demand instantly.

[0630] A "smartphone application" refers to software that runs on a mobile phone terminal and provides specific functions and services to users.

[0631] This invention provides a system that enables brick-and-mortar store operators to monitor sales trends and demand in real time and implement dynamic pricing and personalized marketing strategies. Specific embodiments for implementing this system are described below.

[0632] First, the server accesses the company's database to collect past sales data, external market trend data, and seasonal data, thereby providing the basis for highly accurate demand forecasting.

[0633] This collected data is cleaned and pre-processed on the server. Specifically, missing values ​​in the sales data are imputed using the mean and median, and outliers are removed. At this stage, the data is standardized and integrated with market trend and seasonal data.

[0634] The preprocessed data is used to train a machine learning model for demand forecasting. The server divides the data into training data and validation data, selects a time series forecasting model or neural network model, and begins training. For example, by using an LSTM (long short-term memory) model, a model that can accurately predict the next month's sales demand is built.

[0635] Furthermore, the server uses the newly collected data to forecast future sales trends and demand, and provides these forecast results to users via a dashboard, an interface that allows store operators to visually view the demand forecast results.

[0636] Users can then use these forecasts to implement dynamic pricing, which adjusts product prices in real time based on predicted demand. For example, they can maximize profits by applying discounts to items with high inventory and raising prices for items predicted to be in high demand.

[0637] Furthermore, the server automatically generates personalized marketing strategies based on the prediction results and customer data, which are tailored to specific customer segments and can, for example, send promotional emails for specific products to customers.

[0638] As part of the system, a smartphone application that performs real-time demand forecasting and dynamic pricing will also be provided, allowing store operators to easily check demand forecast results and adjust pricing on the go.

[0639] The hardware and software used are as follows: For hardware, a standard server machine and a smartphone terminal are used. For software, Python, Pandas, NumPy, Scikit-learn, Keras, etc. are used. Using these tools, data collection, cleaning, preprocessing, model training, and prediction can be performed efficiently.

[0640] A concrete example is a scenario in which the system needs to predict the demand for each product this Christmas season based on sales figures from last December, and then make pricing suggestions based on that. The user can request this by entering the following prompt:

[0641] "Based on last December's sales figures, please predict the demand for each product during this year's Christmas season. Also, please make pricing suggestions based on your results."

[0642] In this way, the present invention not only dramatically improves the operational efficiency of physical stores, but also realizes highly accurate demand forecasting and dynamic price adjustments.

[0643] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0644] Step 1:

[0645] A user launches a smartphone application and submits a request to collect historical sales data, external market trend data, and seasonality data.

[0646] Input: Request data

[0647] Output: Collection request

[0648] Specific Operation: When a user presses a button on the application to start data collection, a request is sent to the server to collect historical sales data, external market trend data, and seasonal data.

[0649] Step 2:

[0650] The server accesses the company's database to retrieve historical sales data, collects market trend data from external APIs, and loads pre-defined seasonality data.

[0651] Input: Collection request

[0652] Output: Raw data

[0653] What happens: The server queries the company's database to retrieve historical sales data, and simultaneously sends a request to an external API to retrieve market trend data and load seasonal data files.

[0654] Step 3:

[0655] The server cleans and pre-processes the collected data, imputing missing values, removing outliers, and standardizing the data. It also integrates market trend and seasonal data with sales data.

[0656] Input: Raw data

[0657] Output: Preprocessed data

[0658] What it does: It uses Pandas to impute missing data values ​​with the mean or median, detect and remove outliers, and standardize and combine data into a single dataset.

[0659] Step 4:

[0660] The server uses the preprocessed data to train a machine learning model for demand forecasting. It splits the data into training data and validation data and uses a time series model such as LSTM.

[0661] Input: Preprocessed data

[0662] Output: A trained model

[0663] Specific operation: Using Scikit-learn and Keras, the data is divided into training data and validation data, and an LSTM model is constructed and trained.

[0664] Step 5:

[0665] The server uses the newly collected data to predict future sales trends and demand with a trained model.

[0666] Input: New raw data

[0667] Output: Demand forecast results

[0668] Specific operation: After preprocessing the newly collected sales data and market trend data, it is input into the trained model to predict future demand.

[0669] Step 6:

[0670] The server dynamically sets prices using a pricing algorithm based on the prediction results.

[0671] Input: Demand forecast results

[0672] Output: Pricing data

[0673] Specific operation: Based on the prediction results, the server performs dynamic pricing, lowering the prices of items with high inventory and raising the prices of items with high demand.

[0674] Step 7:

[0675] The forecast results and pricing data are generated as a dashboard and displayed on a smartphone application.

[0676] Input: Forecast results, pricing data

[0677] Output: Dashboard

[0678] Specific operation: The server visualizes the forecast results and pricing data, generates a dashboard, and delivers it to a smartphone application.

[0679] Step 8:

[0680] Users can check the dashboard through a smartphone application and adjust prices and manage inventory in physical stores based on sales strategies.

[0681] Input: Dashboard

[0682] Output: Sales strategy implementation

[0683] Specific Action: User reviews the application dashboard and implements specific sales strategies based on predicted demand and pricing.

[0684] Through this process, physical store operators can achieve highly accurate demand forecasts and quick price adjustments.

[0685] 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.

[0686] As an embodiment of the present invention, we provide a system that uses AI and machine learning to predict future sales trends and demand, and generates marketing strategies based on user emotions using an emotion engine. This system utilizes past sales data, market trend data, seasonal data, and customer emotion data to enable companies to effectively optimize sales plans, inventory management, and marketing strategies.

[0687] First, the server accesses the company's database to collect past sales data, external market trend data, and seasonal data. For example, it obtains detailed information such as monthly sales data for the past five years and the sales volume and price of each product. It also obtains market trend data from an external API. It also accesses a calendar event database to collect data on seasonal events such as Christmas and Valentine's Day.

[0688] The server then cleans and preprocesses the collected data. For example, it imputes missing values ​​in the sales data using the mean or median, and removes outliers. This preprocessing phase also scales the data and integrates sales data, market trend data, and seasonality data. The resulting dataset is then used to train machine learning models.

[0689] The server then uses the preprocessed data to train a machine learning model. Here, the data is split into training data and validation data, and a time series forecasting model or neural network model is selected and trained. For example, the server uses an LSTM (long short-term memory) model to train a model to predict sales demand for the next month. During the training process, the data is repeatedly fed into the model, and patterns are learned.

[0690] The system also incorporates an emotion engine that analyzes user emotions and optimizes marketing strategies based on those emotions. Specifically, the server uses customer data to analyze customer emotions in real time. For example, if a customer expresses positive emotions, the system may introduce new products or make cross-selling suggestions. Conversely, if a customer expresses negative emotions, the system may provide customer support or send discount coupons.

[0691] Once training is complete, the server uses the newly collected data to predict future sales trends and demand. These predictions are stored locally and then made available to users via a dashboard. For example, the dashboard might display sales forecasts for a specific product during the upcoming Christmas season.

[0692] Once the forecast results are available, the server then performs dynamic pricing, adjusting product prices in real time based on predicted demand. For example, discounts are applied to items with high inventory, and prices are increased for items predicted to be in high demand. This dynamic pricing allows businesses to maximize profits.

[0693] Furthermore, the device displays the forecast results to the user through a dashboard. Information is presented using visual elements (graphs, charts) for easy understanding. Users can also monitor the effectiveness of their marketing efforts in real time, allowing them to quickly determine the success of their efforts and any necessary adjustments.

[0694] As described above, the present invention provides a system that provides dynamic pricing based on highly accurate demand forecasts and an optimized marketing strategy using an emotion engine, contributing to corporate sales planning, inventory management, and profit maximization.

[0695] The processing flow will be explained below.

[0696] Step 1:

[0697] The server accesses the company's database and collects past sales data, such as monthly sales data for the past five years and detailed information such as the sales volume and price of each product.

[0698] Step 2:

[0699] The server uses external APIs to collect market trend data, specifically, Google Trends and social media data, to understand consumer interest trends in the market.

[0700] Step 3:

[0701] The server accesses a calendar event database and collects data that reflects seasonality, for example, collecting data about seasonal events such as Christmas and Valentine's Day.

[0702] Step 4:

[0703] The server cleans the collected data by detecting missing values ​​in the dataset, imputing them with the mean or median, and removing outliers.

[0704] Step 5:

[0705] The server pre-processes the cleaned data, including scaling and normalizing it, and integrating sales data, market trend data, and seasonality data.

[0706] Step 6:

[0707] The server splits the preprocessed dataset into training and validation data, typically 80% of the total data for training and 20% for validation.

[0708] Step 7:

[0709] The server selects a time series prediction model or neural network model, for example, a long short-term memory (LSTM) model.

[0710] Step 8:

[0711] The server uses the training data to train the machine learning model by repeatedly inputting the data into the model and letting it learn patterns.

[0712] Step 9:

[0713] The server evaluates the performance of the trained model using validation data, using metrics such as MSE (Mean Squared Error) and MAE (Mean Absolute Error).

[0714] Step 10:

[0715] The server uses the new data to predict future sales trends and demand, for example, predicting the sales volume of a particular product for the next month.

[0716] Step 11:

[0717] The server stores the prediction results in a database and generates a dashboard, which displays the prediction results in the form of graphs and charts.

[0718] Step 12:

[0719] The terminal displays the prediction results to the user through a dashboard, providing visual information that is easy for the user to understand.

[0720] Step 13:

[0721] The server dynamically sets prices based on the demand forecast results, for example, raising prices for products with high predicted demand and discounting products with high inventory.

[0722] Step 14:

[0723] The user implements the individually optimized marketing strategy provided by the server, specifically, sending personalized promotional emails to specific customer segments.

[0724] Step 15:

[0725] The server runs an emotion engine using customer data, such as past purchase history, browsing history, and social media comments, to analyze customer emotions in real time.

[0726] Step 16:

[0727] The server uses the emotion engine to generate more effective marketing strategies based on the obtained emotion data, such as introducing new products to customers who show positive emotions and providing support or discount coupons to customers who show negative emotions.

[0728] Step 17:

[0729] Users can monitor the effectiveness of their marketing initiatives in real time, allowing them to quickly determine the success of their initiatives and make any necessary adjustments.

[0730] Example 2

[0731] Next, a description will be given of 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."

[0732] For businesses, accurate demand forecasting and determining effective marketing strategies are important issues that directly affect sales growth and inventory management. However, many businesses are unable to properly utilize past data, making it difficult to accurately reflect seasonality and market trends. Formulating appropriate marketing strategies based on customer sentiment is also a challenge. The present invention aims to solve these problems and provide a set of tools that enable businesses to optimize sales planning, inventory management, and marketing strategies.

[0733] 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.

[0734] In this invention, the server includes means for collecting past transaction data, external market trend data, and seasonal information, means for cleaning and pre-processing the collected data, means for training a machine learning model for demand forecasting using the pre-processed data, means for forecasting future sales trends and demand using the trained model, means for generating a visualization interface for displaying the forecast results, means for dynamic pricing based on the forecast results, and means for analyzing customer sentiment data using a sentiment analyzer and generating individually optimized marketing strategies based on the results, thereby enabling highly accurate sales forecasting and optimization of marketing strategies based on customer sentiment.

[0735] "Past transaction data" refers to information about all sales activities conducted by a company in the past, including details such as sales quantity, sales price, and sales date and time.

[0736] "External market trend data" refers to information that shows overall market trends, competitor activities, consumer purchasing trends, etc., and is obtained from external information providers or APIs.

[0737] "Seasonal information" is data that indicates fluctuations in sales related to specific seasons or events, and includes sales trends during specific periods such as Christmas and Valentine's Day.

[0738] "Cleaning" refers to the process of removing errors, missing values, and inappropriate data from collected data to improve the quality of the data.

[0739] "Preprocessing" is the process of preparing data in a format suitable for a machine learning model, and includes scaling and standardizing the data, and filling in missing values.

[0740] A "machine learning model" is an algorithm or system used to make predictions, classifications, and optimizations based on input data, and includes LSTM and neural networks.

[0741] "Training" refers to the process of using existing data to teach a machine learning model so that it can make appropriate predictions and recommendations for new data.

[0742] "Future Sales Trends" refers to future sales trends and changes in demand predicted by the machine learning model.

[0743] "Prediction results" refer to the results of a machine learning model's predictions about the future, and include information such as specific sales volumes and demand patterns.

[0744] A "visualization interface" is a means for displaying prediction results in a visual format such as a graph or chart, making it easy for users to understand.

[0745] "Dynamic pricing" refers to the practice of adjusting product prices in real time based on predicted demand, raising or lowering prices based on inventory levels and supply-demand balances.

[0746] An "emotion analyzer" is a device or software that analyzes customer emotions and determines the emotional state of customers by analyzing reviews, social media posts, etc.

[0747] An "individually optimized marketing strategy" refers to a strategy that designs and executes optimal marketing methods for individual customers based on collected customer data and the results of sentiment analysis.

[0748] This invention is a system that uses AI and machine learning to predict future sales trends and demand, and further uses an emotion engine to generate marketing strategies based on user emotions. This system aims to help companies optimize their sales plans, inventory management, and marketing strategies.

[0749] Data collection

[0750] server

[0751] The server first accesses the company's database to retrieve transaction data from the past five years. This data includes details such as the sales quantity, sales price, and sales date for each product. It also collects market trend data through an external API. This market trend data includes consumer purchasing trends and competitor activity. The server also accesses a calendar event database to collect information about seasonal events such as Christmas and Valentine's Day.

[0752] Data Preprocessing

[0753] server

[0754] The server cleans and preprocesses the collected data. Specifically, if there are missing values ​​in the sales data, they are imputed using the mean or median. Outliers are also removed as soon as they are detected. The data is then scaled and converted into a unified format. Sales data, market trend data, and seasonal data are integrated into a single dataset.

[0755] Training a machine learning model

[0756] server

[0757] The preprocessed dataset is used to train the machine learning model. The dataset is divided into training data and validation data. The server selects an LSTM (long short-term memory) model and trains it to learn sales demand by inputting data into the model. This process is performed so that the model can learn sales patterns and improve its prediction accuracy for future data.

[0758] Predicting future sales trends

[0759] server

[0760] Once training is complete, the newly collected data is used to predict future sales trends and demand, such as predicting sales volumes for a particular product during the upcoming Christmas season. These predictions are stored on the server and later provided to users via a dashboard.

[0761] Utilizing the Emotion Engine

[0762] server

[0763] The server uses a sentiment analyzer to analyze user sentiment in real time. For example, it analyzes customer reviews and social media posts, and if the sentiment is positive, it introduces new products or suggests cross-selling. Conversely, if the sentiment is negative, it provides customer support or sends discount coupons.

[0764] Dynamic Pricing

[0765] server

[0766] Based on the predictions, the server dynamically sets prices, for example, applying discounts to items with high inventory and raising prices for items predicted to be in high demand, aiming to maximize the company's profits.

[0767] Viewing the Dashboard

[0768] Terminal

[0769] The dashboard displays the forecast results sent from the server, using visual elements such as graphs and charts to provide information that users can easily understand. Users can also monitor the effectiveness of their marketing efforts in real time, quickly determining the success of their efforts and any necessary adjustments.

[0770] Specific examples

[0771] For example, the server collects data for the Christmas shopping season in December. First, it collects sales data from past Christmas seasons from the database, then obtains market trend data using an external API. It also collects Christmas-related data from the calendar event database. The server then imputes any missing values ​​in the collected data using the average value, scales and integrates all the data. Once all the datasets are ready, it uses them to train an LSTM model. It then performs predictions on the newly acquired data and displays the results on a dashboard. Finally, it uses an emotion engine to introduce new products if customers have positive emotions.

[0772] Prompt Sentence Examples

[0773] "For the Christmas sales season leading up to December, collect Christmas season sales data, market trend data, and seasonal event data from the past five years, and use this data to make predictions using an LSTM model and display them on a dashboard. In addition, analyze customer sentiment data in real time, and if customers show positive sentiment, introduce new products or make cross-selling suggestions."

[0774] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0775] Step 1: Data collection

[0776] server

[0777] The server first accesses the company's database and retrieves transaction data from the past five years, including details such as the quantity, price, and date of each product sold.

[0778] Input: Company database

[0779] Output: Transaction data for the past 5 years

[0780] The server then collects market trend data through external APIs, including consumer purchasing habits and competitor activity.

[0781] Input: External API

[0782] Output: Market trend data

[0783] Additionally, the server accesses a calendar events database to collect information about seasonal events such as Christmas and Valentine's Day.

[0784] Input: Calendar event database

[0785] Output: Seasonal event data

[0786] Step 2: Data Preprocessing

[0787] server

[0788] The server cleans and preprocesses all collected data. Specifically, if there are missing values ​​in the sales data, they are imputed using the mean or median. Outliers are also removed as soon as they are detected. The server also scales the data and converts it into a unified format. Finally, the sales data, market price trend data, and seasonal data are integrated into a single dataset.

[0789] Inputs: trade data, market trend data, seasonal event data

[0790] Output: Preprocessed dataset

[0791] Step 3: Train the model

[0792] server

[0793] The preprocessed dataset is used to train a machine learning model. The dataset is divided into training data and validation data. The server selects an LSTM (long short-term memory) model and inputs the data multiple times to train the model on sales demand. This training process allows the model to learn sales patterns and improve its prediction accuracy for future data.

[0794] Input: Preprocessed dataset

[0795] Output: A trained LSTM model

[0796] Step 4: Predict future sales trends

[0797] server

[0798] Using the trained LSTM model, the server analyzes newly collected data and predicts future sales trends and demand. For example, it predicts the sales volume of a particular product during the upcoming Christmas season. The prediction results are stored on the server and later provided to users via a dashboard.

[0799] Input: Newly collected data, trained LSTM model

[0800] Output: Future sales trend forecast results

[0801] Step 5: Leverage the Emotion Engine

[0802] server

[0803] The server uses a sentiment analyzer to analyze user emotions in real time. Specifically, it analyzes customer reviews and social media posts, and introduces new products and makes cross-selling suggestions to customers who show positive emotions. Conversely, it provides customer support and sends discount coupons to customers who show negative emotions.

[0804] Input: Customer reviews, social media posts

[0805] Output: Emotion-based marketing strategies

[0806] Step 6: Dynamic Pricing

[0807] server

[0808] Based on the predictions, the server dynamically sets prices, applying discounts to items with high inventory and raising prices for items predicted to be in high demand, thereby maximizing profits for the company.

[0809] Input: Future sales trend forecast results

[0810] Output: Real-time adjusted product prices

[0811] Step 7: View the dashboard

[0812] Terminal

[0813] The dashboard displays the prediction results sent from the server. The dashboard uses visual elements (graphs, charts) to provide users with easy-to-understand information. Users can also monitor the effectiveness of their marketing initiatives in real time through the dashboard, allowing them to quickly determine the success of their initiatives and any necessary adjustments.

[0814] Input: Future sales trend prediction results, emotion-based marketing strategies

[0815] Output: Visually displayed forecast results and strategies

[0816] (Application example 2)

[0817] Next, a description will be given of Application 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."

[0818] Modern marketing strategies and sales plans must respond quickly and accurately to diverse customer needs. However, traditional systems are based solely on past sales data and market trend data, which means they cannot fully consider customer sentiment or real-time conditions. Other issues include the difficulty of improving the accuracy of demand forecasts and dynamic pricing, and providing individually optimized marketing strategies in real time. This makes it difficult for companies to optimize sales plans and inventory management, leading to a demand for systems that contribute to maximizing profits.

[0819] 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.

[0820] In this invention, the server includes means for collecting past sales data, external market trend data, and seasonal data, means for cleaning and pre-processing the collected data, means for training a machine learning model for demand forecasting using the pre-processed data, means for collecting and analyzing customer sentiment data, and means for optimizing marketing strategies based on customer sentiment, thereby improving the accuracy of demand forecasting and marketing strategies and enabling optimal sales planning and inventory management that take customer sentiment into account.

[0821] "Sales data" refers to historical data relating to past sales, inventory information, product sales quantities and prices, and the like.

[0822] "Market trend data" refers to data on customer purchasing trends, popular product trends, and economic conditions in a broad range of markets.

[0823] "Seasonal data" refers to data on fluctuations in customer purchasing behavior associated with specific seasons or events, such as Christmas or Valentine's Day.

[0824] A "machine learning model" is a collection of algorithms that learn patterns from data and make predictions or classifications based on those patterns.

[0825] "Emotional data" refers to data on the emotional movements and feelings of customers that can be obtained from their facial expressions, voice, and behavior.

[0826] A "marketing strategy" is a plan for a series of advertising, promotion, pricing, and other activities designed to achieve a specific objective.

[0827] "Training data" is a learning dataset used when building a machine learning model.

[0828] "Validation data" is a dataset used to evaluate the performance of a trained machine learning model.

[0829] A "dashboard" is an interface that visually displays forecast results and other important information in a format that is easy for users to understand.

[0830] "Dynamic pricing" is a method of adjusting product prices in real time based on demand forecasts and inventory status.

[0831] "Demand forecasting" refers to predicting future customer purchasing behavior and sales volumes based on past data and trend information.

[0832] As an embodiment of the present invention, a server uses AI and machine learning to predict future sales trends and demand, and also uses an emotion engine to generate a marketing strategy based on user emotions. Specific implementation means are described below.

[0833] Data collection

[0834] The server accesses the company's database to collect past sales data, external market trend data, and seasonal data, including the "sales data," "market trend data," and "seasonal data" previously specified. For example, the server uses an API to obtain external trend data and retrieves past sales history from the company's internal database.

[0835] Data Cleaning and Preprocessing

[0836] The collected data is cleaned, missing values ​​are filled, and outliers are removed. This is a necessary step to increase the reliability of the data. For example, missing values ​​are filled with the mean or median. The data is then scaled and sales data, market trend data, and seasonal data are integrated.

[0837] Training a machine learning model

[0838] The preprocessed data is used to train a machine learning model. The model used here is a "long short-term memory (LSTM) model" or a "neural network model." The server splits the data into training data and validation data and performs training.

[0839] Demand forecasting

[0840] The trained model is used to predict future sales trends and demand, for example, predicting demand for specific products around seasonal events like Christmas and Valentine's Day. These predictions are then displayed on a dashboard as graphs and charts that users can easily understand.

[0841] Emotion data collection and analysis

[0842] The server collects customer emotional data from the smartphone's camera and microphone. Using specialized software for emotion analysis, it analyzes facial expressions and voice in real time. This emotional data is used to optimize marketing strategies.

[0843] Optimizing your marketing strategy

[0844] Optimize your marketing strategies based on customer sentiment data. For example, if a customer expresses positive sentiment, introduce new products or offer cross-selling suggestions. If a customer expresses negative sentiment, offer customer support or send discount coupons.

[0845] Dynamic Pricing

[0846] Dynamically adjust pricing based on demand forecasts. If demand for a particular product is predicted to increase, prices are raised, while discounts are offered if inventory is high. This allows for real-time price adjustments to maximize profits.

[0847] Dashboard View

[0848] It generates a dashboard that visually displays forecast results, sentiment analysis results, and pricing information, allowing users to easily check upcoming sales plans and marketing initiatives and quickly adjust them as needed.

[0849] For example, the following prompt sentence is input to the generative AI model:

[0850] "Please forecast the demand for home video game consoles during the next Christmas season."

[0851] "Analyze recent user reviews to identify products that have received positive feedback."

[0852] This allows companies to develop highly accurate sales strategies based on market trends and customer sentiment, and implement optimized marketing strategies.

[0853] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0854] Step 1:

[0855] The server collects historical sales data from the company's database and obtains external market trend data and seasonal data through API. The input is database connection information and API key, and the output is the integrated raw data set. Specifically, it issues database queries and API requests and stores the obtained data in local storage.

[0856] Step 2:

[0857] The server performs data cleaning on the collected raw dataset. The input is the integrated raw dataset, and the output is a clean dataset. Specific operations include imputing missing values ​​(calculating the mean and median), removing outliers, and performing data scaling. This results in a dataset suitable for training machine learning models.

[0858] Step 3:

[0859] The server trains a machine learning model using a clean dataset. The input is the clean dataset, and the output is the trained model. Specifically, the data is split into training data and validation data, and training is performed using an LSTM model or a neural network model. Machine learning libraries such as TensorFlow and Keras are used to train the model.

[0860] Step 4:

[0861] The server uses the trained model to predict future sales trends and demand. The input is a newly collected dataset, and the output is the prediction result. Specifically, new data is input into the model, and predictions are made to calculate demand fluctuations and sales trends.

[0862] Step 5:

[0863] The device collects user emotional data through the smartphone's camera and microphone. The input is video and audio data, and the output is the result of emotional analysis. Specifically, it uses an emotion analysis engine to analyze changes in facial expressions and voice tone in real time to extract emotional data.

[0864] Step 6:

[0865] The server generates an individually optimized marketing strategy based on the emotional data. The input is the result of the emotional analysis, and the output is a set of marketing measures. Specifically, it runs an algorithm that performs promotions and product recommendations according to the emotional state.

[0866] Step 7:

[0867] The server performs dynamic pricing based on the demand forecast results. The inputs are the demand forecast results and inventory data, and the output is adjusted price information. Specifically, it runs an algorithm that raises prices when demand is high and applies discounts when inventory is high.

[0868] Step 8:

[0869] The terminal generates a dashboard and displays the forecast results, sentiment analysis results, and pricing information. The input is the forecast results and sentiment analysis results, and the output is a visually displayed dashboard. Specific operations include displaying information using graphs and charts, allowing users to easily check sales forecasts and marketing strategies.

[0870] Through these steps, the server and terminals work together to provide a system that realizes highly accurate demand forecasting, dynamic pricing, and emotion-based marketing strategies.

[0871] 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.

[0872] 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> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) 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.

[0873] 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.

[0874] [Third embodiment]

[0875] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0876] 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.

[0877] 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).

[0878] 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.

[0879] 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.

[0880] 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).

[0881] 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.

[0882] 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.

[0883] 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.

[0884] 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.

[0885] 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.

[0886] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0887] As an embodiment of the present invention, we provide a system that uses AI and machine learning to forecast future sales trends and demand. This system aims to enable companies to effectively optimize sales planning, inventory management, and marketing strategies by using past sales data, market trend data, and seasonal data.

[0888] The specific program processing of this system is as follows.

[0889] First, the server accesses the company's database to collect past sales data, external market trend data, and seasonal data. For example, the server obtains monthly sales data for the past five years and then obtains market trend data from an external API. In this way, multifaceted data collection forms the basis for a more accurate forecasting model.

[0890] The server then cleans and preprocesses the collected data. For example, missing values ​​in the sales data are imputed using the mean or median. Outliers are also removed. This preprocessing phase normalizes the data and integrates market trend and seasonal data with the sales data. The resulting dataset is then used to train the machine learning model.

[0891] The server then uses the preprocessed data to train a machine learning model. Here, the data is split into training data and validation data, and a time series forecasting model or neural network model is selected and trained. For example, the server uses an LSTM (long short-term memory) model to train a model to predict sales demand for the next month. During this training process, the data is repeatedly fed into the model, and patterns are learned.

[0892] Once training is complete, the server uses the newly collected data to predict future sales trends and demand. These predictions are stored locally and then made available to users via a dashboard. For example, the dashboard might display sales forecasts for a specific product during the upcoming Christmas season.

[0893] Once the forecast results are available, the server then performs dynamic pricing, adjusting product prices in real time based on predicted demand. For example, discounts are applied to items with high inventory, and prices are increased for items predicted to be in high demand. This dynamic pricing allows businesses to maximize profits.

[0894] Furthermore, the user implements individually optimized marketing strategies provided by the server. The server generates personalized promotional emails for specific customer segments based on the prediction results and customer data. The user can then run advertising campaigns based on this and monitor the results in real time.

[0895] In this way, the present invention provides dynamic pricing and optimized marketing strategies based on highly accurate demand forecasts, realizing a system that contributes to corporate sales planning, inventory management, and profit maximization.

[0896] The processing flow will be explained below.

[0897] Step 1:

[0898] The server accesses the company's database and collects past sales data, such as monthly sales data for the past five years and detailed information such as the sales volume and price of each product.

[0899] Step 2:

[0900] The server uses external APIs to collect relevant market trend data, specifically Google Trends and social media data, to understand consumer interest trends in the market.

[0901] Step 3:

[0902] The server accesses a calendar event database to retrieve data that reflects seasonality, for example, collecting data about seasonal events such as Christmas and Valentine's Day.

[0903] Step 4:

[0904] The server cleans the collected data by detecting missing values ​​in the dataset, imputing them with the mean or median, and removing outliers.

[0905] Step 5:

[0906] The server pre-processes the cleaned data, including scaling and normalizing the data, and integrating sales data, market trend data, and seasonality data.

[0907] Step 6:

[0908] The server splits the preprocessed dataset into training and validation data, typically 80% of the total data for training and 20% for validation.

[0909] Step 7:

[0910] The server selects a time series prediction model or neural network model, for example, a long short-term memory (LSTM) model.

[0911] Step 8:

[0912] The server uses the training data to train the machine learning model, which involves repeatedly feeding the data into the model and letting it learn patterns.

[0913] Step 9:

[0914] The server evaluates the performance of the trained model using validation data, using metrics such as MSE (Mean Squared Error) and MAE (Mean Absolute Error).

[0915] Step 10:

[0916] The server uses the new data to predict future sales trends and demand, for example, predicting the sales volume of a particular product for the next month.

[0917] Step 11:

[0918] The server stores the prediction results in a database and generates a dashboard, which displays the prediction results in the form of graphs and charts.

[0919] Step 12:

[0920] The device displays the prediction results to the user through a dashboard, providing visual information that is easy for the user to understand.

[0921] Step 13:

[0922] The server dynamically sets prices based on the demand forecast results, for example, raising prices for products with high predicted demand and discounting products with high inventory.

[0923] Step 14:

[0924] The user implements individually optimized marketing strategies generated by the server, for example, sending personalized promotional emails to specific customer segments.

[0925] Step 15:

[0926] Users can monitor the effectiveness of their marketing efforts in real time, allowing them to quickly determine the success of their efforts and make any necessary adjustments.

[0927] Example 1

[0928] Next, a description will be given of 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."

[0929] Conventional sales forecasting systems often base their forecasts solely on past sales data and are unable to comprehensively incorporate external market trend data or seasonal data, resulting in low forecast accuracy. Furthermore, they lack effective means for implementing dynamic pricing and individually optimized marketing strategies. A system that can solve this issue and provide highly accurate demand forecasts and dynamic pricing and marketing strategies based on those forecasts is needed.

[0930] 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.

[0931] In this invention, the server includes means for collecting past sales data, external market trend data, and seasonal data, means for cleaning and pre-processing the collected data, means for training a machine learning model for demand forecasting using the pre-processed data, means for forecasting future sales trends and demand using the trained model, means for generating a recording medium for displaying the forecast results, means for dynamic pricing based on the forecast results, and means for generating an individually optimized marketing strategy based on customer data, thereby enabling highly accurate demand forecasting, dynamic pricing, and the implementation of individually optimized marketing strategies.

[0932] "Sales data" is a collection of information about sales performance of a particular product over a past period.

[0933] "Market trend data" is a collection of information that indicates the current situation and future trends in a particular market or industry.

[0934] "Seasonal data" is a collection of information that indicates sales trends and demand fluctuations associated with particular seasons or annual events.

[0935] "Data cleaning" is the process of removing inaccurate, incomplete, or unnecessary values ​​from collected data to improve data quality.

[0936] "Preprocessing" refers to a series of steps taken to convert collected data into a format suitable for analysis and machine learning.

[0937] A "machine learning model" is an algorithm or mathematical model that learns patterns from data and performs tasks such as prediction and classification.

[0938] "Training" is the process of providing a machine learning model with large amounts of data and having it learn from that data to improve its prediction accuracy.

[0939] A "recording medium" is a physical or electronic device or means for storing and displaying digital data.

[0940] "Dynamic pricing" is a method of adjusting prices in real time based on demand forecasts and market conditions.

[0941] A "marketing strategy" is a series of plans and measures implemented to promote sales and increase brand awareness among specific customer segments.

[0942] This invention relates to a sales forecasting system, specifically a system that uses AI and machine learning technologies to forecast future sales trends and demand. This system is designed to help companies more effectively plan and execute sales plans, inventory management, and marketing strategies by utilizing past sales data, market trend data, and seasonal data.

[0943] Hardware and software used

[0944] 1. Server: Manages all data collection, preprocessing, machine learning model training, and dynamic pricing using libraries such as Python, Pandas, Scikit-learn, TensorFlow, or PyTorch.

[0945] 2. Database: Stores and manages company sales data and external trend data. Specifically, MySQL or PostgreSQL can be used.

[0946] 3. External API: Used to obtain market trend data.

[0947] 4. Web dashboard: An interface for providing predictions and analysis results to users. Specifically, JavaScript, D3.js, or Tableau may be used.

[0948] Data collection and preprocessing

[0949] The server accesses the company's database to collect past sales data, external market trend data, and seasonal data. This data allows for more accurate forecasts based on multifaceted information. Specifically, the server periodically sends API requests to collect and store the necessary data.

[0950] Next, the server cleans and preprocesses the data. This step includes imputing missing values, removing outliers, and normalizing the data. For example, it uses the Pandas fillna() method to impute missing values ​​with the mean.

[0951] Training a machine learning model

[0952] The server uses the preprocessed data to train a machine learning model. Here, the data is divided into training data and validation data, and a long short-term memory (LSTM) model is used to perform time series prediction. Specifically, an LSTM model is built and trained using TensorFlow.

[0953] Future sales trends and demand forecasts

[0954] Using the trained model, the server predicts future sales trends and demand based on newly collected data. These predictions are stored in a database and made available to users through a web dashboard. For example, the dashboard might predict that demand for a particular product will increase during the upcoming Christmas season.

[0955] Dynamic Pricing

[0956] The server then dynamically sets prices based on the forecast, for example, discounting items with high inventory and raising prices for items predicted to be in high demand. This dynamic pricing is a key factor in helping companies maximize profits.

[0957] Individually optimized marketing strategies

[0958] Based on the prediction results and customer data provided by the server, users can create personalized promotional emails and advertising campaigns for specific customer segments. For example, emails can be automatically generated to strongly appeal to specific customer demographics about a specific product.

[0959] Prompt Sentence Examples

[0960] "Please provide us with your sales forecast for electronic products for the upcoming Christmas season. Also, please identify which products will sell best and suggest a corresponding marketing strategy."

[0961] In this way, this system provides dynamic pricing and optimized marketing strategies based on highly accurate demand forecasts, contributing to corporate sales planning, inventory management, and profit maximization.

[0962] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0963] Step 1:

[0964] Data collection

[0965] The server accesses the company's database to collect historical sales data, external market trend data, and seasonality data.

[0966] Specific behavior:

[0967] The server periodically executes SQL queries to extract sales data from the company's database. The server sends HTTP requests to retrieve market trend data from an external API. The server stores this data in a local database.

[0968] input:

[0969] Corporate database connection information, external API endpoints.

[0970] output:

[0971] Collection of historical sales data, market trend data, and seasonality data.

[0972] Step 2:

[0973] Data Cleaning and Preprocessing

[0974] The server cleans the collected data, filling in missing values, removing outliers, and normalizing it.

[0975] Specific behavior:

[0976] The server uses the Pandas library to load the data into a data frame, uses the fillna() method to impute missing values, boxplots and Z-scores to detect and remove outliers, and normalizes each data set to create a unified dataset.

[0977] input:

[0978] Collected sales data, market trend data, and seasonality data.

[0979] output:

[0980] Cleaned and preprocessed integrated dataset.

[0981] Step 3:

[0982] Data partitioning

[0983] The server splits the combined dataset into training data and validation data.

[0984] Specific behavior:

[0985] The server splits the data into training and validation data using Scikit-learn's train_test_split function.

[0986] input:

[0987] Cleaned and preprocessed integrated dataset.

[0988] output:

[0989] Training and validation data.

[0990] Step 4:

[0991] Training a machine learning model

[0992] The server trains the LSTM model using the training data.

[0993] Specific behavior:

[0994] The server uses the TensorFlow library to build an LSTM model and trains it using the model.fit() method. During the training process, the validation data is used to evaluate the accuracy of the model.

[0995] input:

[0996] Training data, validation data.

[0997] output:

[0998] A trained LSTM model.

[0999] Step 5:

[1000] Future predictions

[1001] The server uses the trained model to predict future sales trends and demand.

[1002] Specific behavior:

[1003] The server preprocesses the newly collected data and inputs it into the trained LSTM model, using the model.predict() method to predict the sales demand for the next month.

[1004] input:

[1005] Newly collected data, trained LSTM model.

[1006] output:

[1007] Sales trends and demand forecast results.

[1008] Step 6:

[1009] Saving and displaying prediction results

[1010] The server stores the prediction results in a database and displays them on a dashboard that users can access.

[1011] Specific behavior:

[1012] The server stores the prediction results in a database and provides them to users via a web dashboard, where users can check the prediction results.

[1013] input:

[1014] Sales trends and demand forecast results.

[1015] output:

[1016] Prediction results displayed on a dashboard.

[1017] Step 7:

[1018] Dynamic Pricing

[1019] The server dynamically sets prices based on the prediction results.

[1020] Specific behavior:

[1021] The server runs a price optimization algorithm and adjusts prices based on inventory and demand forecasts, then feeds the adjusted prices into the company's ERP system.

[1022] input:

[1023] Sales trends and demand forecast results, inventory data.

[1024] output:

[1025] Dynamically adjusted pricing information.

[1026] Step 8:

[1027] Implementing marketing strategies

[1028] The user executes an individually optimized marketing campaign based on the marketing strategy provided by the server.

[1029] Specific behavior:

[1030] The server automatically generates promotional emails and advertising campaigns and sends them to specific customer segments, and users can monitor the effectiveness of the campaigns in real time using a dashboard.

[1031] input:

[1032] Prediction results, customer data.

[1033] output:

[1034] Individually optimized marketing strategies and execution results.

[1035] These are the specific processing steps of the program for this system, which enables companies to achieve highly accurate demand forecasting, dynamic pricing, and individually optimized marketing strategies.

[1036] (Application example 1)

[1037] Next, a description will be given of Application 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."

[1038] Traditional sales trend and demand forecasting systems often lack sufficient forecasting accuracy, making them ineffective for companies' inventory management and marketing strategies. Furthermore, these systems require large desktop environments and servers and lack real-time responsiveness. As a result, it is difficult for brick-and-mortar store operators to respond quickly to changing market conditions and special seasonal events. Therefore, there is a need for a system that can achieve more accurate forecasting and dynamic pricing, thereby improving the operational efficiency of brick-and-mortar stores.

[1039] 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.

[1040] In this invention, the server includes means for collecting past sales data, external market trend data, and seasonal data, means for cleaning and preprocessing the collected data, means for training a machine learning model for demand forecasting using the preprocessed data, means for forecasting future sales trends and demand using the trained model, means for generating a dashboard for displaying the forecast results, means for dynamic pricing based on the forecast results, means for generating individually optimized marketing strategies based on customer data, and means for providing a smartphone application that performs real-time demand forecasting and dynamic pricing. This enables physical store operators to perform highly accurate demand forecasting and rapid price adaptation, thereby optimizing inventory management and marketing strategies.

[1041] "Past sales data" refers to numerical data and records relating to sales that have occurred in the past.

[1042] "External market trend data" refers to data obtained from external sources that indicates market trends and tendencies.

[1043] "Seasonal data" refers to data that indicates seasonal fluctuations and includes information such as consumer behavior at specific times.

[1044] "Means of collection" refers to the methods, tools, and processes used to obtain the necessary data.

[1045] "Cleaning" is a data preparation process that involves tasks such as deleting unnecessary elements and filling in missing values.

[1046] "Preprocessing means" refers to methods and tools used to prepare data in a format suitable for machine learning models.

[1047] "Demand forecasting" refers to estimating the future number of products that will be needed and providing information that can be incorporated into sales plans.

[1048] "Means for training a machine learning model" means the method or process of running a machine learning algorithm on collected data to build the model needed to make a prediction.

[1049] A "trained model" refers to a predictive model that has been learned by a machine learning algorithm using historical data.

[1050] "Means for generating dashboards" refers to methods for creating interfaces and tools that allow users to visually examine data.

[1051] "Dynamic pricing" is the process of adjusting prices in real time based on demand and other factors.

[1052] "Customer data" refers to information about individual customers, including purchase history and behavioral data.

[1053] An "individually optimized marketing strategy" refers to a sales promotion strategy that is customized to effectively reach a specific customer segment.

[1054] "Real-time demand forecasting" refers to systems and tools that can process data in real time and forecast demand instantly.

[1055] A "smartphone application" refers to software that runs on a mobile phone terminal and provides specific functions and services to users.

[1056] This invention provides a system that enables brick-and-mortar store operators to monitor sales trends and demand in real time and implement dynamic pricing and personalized marketing strategies. Specific embodiments for implementing this system are described below.

[1057] First, the server accesses the company's database to collect past sales data, external market trend data, and seasonal data, thereby providing the basis for highly accurate demand forecasting.

[1058] This collected data is cleaned and pre-processed on the server. Specifically, missing values ​​in the sales data are imputed using the mean and median, and outliers are removed. At this stage, the data is standardized and integrated with market trend and seasonal data.

[1059] The preprocessed data is used to train a machine learning model for demand forecasting. The server divides the data into training data and validation data, selects a time series forecasting model or neural network model, and begins training. For example, by using an LSTM (long short-term memory) model, a model that can accurately predict the next month's sales demand is built.

[1060] Furthermore, the server uses the newly collected data to forecast future sales trends and demand, and provides these forecast results to users via a dashboard, an interface that allows store operators to visually view the demand forecast results.

[1061] Users can then use these forecasts to implement dynamic pricing, which adjusts product prices in real time based on predicted demand. For example, they can maximize profits by applying discounts to items with high inventory and raising prices for items predicted to be in high demand.

[1062] Furthermore, the server automatically generates personalized marketing strategies based on the prediction results and customer data, which are tailored to specific customer segments and can, for example, send promotional emails for specific products to customers.

[1063] As part of the system, a smartphone application that performs real-time demand forecasting and dynamic pricing will also be provided, allowing store operators to easily check demand forecast results and adjust pricing on the go.

[1064] The hardware and software used are as follows: For hardware, a standard server machine and a smartphone terminal are used. For software, Python, Pandas, NumPy, Scikit-learn, Keras, etc. are used. Using these tools, data collection, cleaning, preprocessing, model training, and prediction can be performed efficiently.

[1065] A concrete example is a scenario in which the system needs to predict the demand for each product this Christmas season based on sales figures from last December, and then make pricing suggestions based on that. The user can request this by entering the following prompt:

[1066] "Based on last December's sales figures, please predict the demand for each product during this year's Christmas season. Also, please make pricing suggestions based on your results."

[1067] In this way, the present invention not only dramatically improves the operational efficiency of physical stores, but also realizes highly accurate demand forecasting and dynamic price adjustments.

[1068] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1069] Step 1:

[1070] A user launches a smartphone application and submits a request to collect historical sales data, external market trend data, and seasonality data.

[1071] Input: Request data

[1072] Output: Collection request

[1073] Specific Operation: When a user presses a button on the application to start data collection, a request is sent to the server to collect historical sales data, external market trend data, and seasonal data.

[1074] Step 2:

[1075] The server accesses the company's database to retrieve historical sales data, collects market trend data from external APIs, and loads pre-defined seasonality data.

[1076] Input: Collection request

[1077] Output: Raw data

[1078] What happens: The server queries the company's database to retrieve historical sales data, and simultaneously sends a request to an external API to retrieve market trend data and load seasonal data files.

[1079] Step 3:

[1080] The server cleans and pre-processes the collected data, imputing missing values, removing outliers, and standardizing the data. It also integrates market trend and seasonal data with sales data.

[1081] Input: Raw data

[1082] Output: Preprocessed data

[1083] What it does: It uses Pandas to impute missing data values ​​with the mean or median, detect and remove outliers, and standardize and combine data into a single dataset.

[1084] Step 4:

[1085] The server uses the preprocessed data to train a machine learning model for demand forecasting. It splits the data into training data and validation data and uses a time series model such as LSTM.

[1086] Input: Preprocessed data

[1087] Output: A trained model

[1088] Specific operation: Using Scikit-learn and Keras, the data is divided into training data and validation data, and an LSTM model is constructed and trained.

[1089] Step 5:

[1090] The server uses the newly collected data to predict future sales trends and demand with a trained model.

[1091] Input: New raw data

[1092] Output: Demand forecast results

[1093] Specific operation: After preprocessing the newly collected sales data and market trend data, it is input into the trained model to predict future demand.

[1094] Step 6:

[1095] The server dynamically sets prices using a pricing algorithm based on the prediction results.

[1096] Input: Demand forecast results

[1097] Output: Pricing data

[1098] Specific operation: Based on the prediction results, the server performs dynamic pricing, lowering the prices of items with high inventory and raising the prices of items with high demand.

[1099] Step 7:

[1100] The forecast results and pricing data are generated as a dashboard and displayed on a smartphone application.

[1101] Input: Forecast results, pricing data

[1102] Output: Dashboard

[1103] Specific operation: The server visualizes the forecast results and pricing data, generates a dashboard, and delivers it to a smartphone application.

[1104] Step 8:

[1105] Users can check the dashboard through a smartphone application and adjust prices and manage inventory in physical stores based on sales strategies.

[1106] Input: Dashboard

[1107] Output: Sales strategy implementation

[1108] Specific Action: User reviews the application dashboard and implements specific sales strategies based on predicted demand and pricing.

[1109] Through this process, physical store operators can achieve highly accurate demand forecasts and quick price adjustments.

[1110] 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.

[1111] As an embodiment of the present invention, we provide a system that uses AI and machine learning to predict future sales trends and demand, and generates marketing strategies based on user emotions using an emotion engine. This system utilizes past sales data, market trend data, seasonal data, and customer emotion data to enable companies to effectively optimize sales plans, inventory management, and marketing strategies.

[1112] First, the server accesses the company's database to collect past sales data, external market trend data, and seasonal data. For example, it obtains detailed information such as monthly sales data for the past five years and the sales volume and price of each product. It also obtains market trend data from an external API. It also accesses a calendar event database to collect data on seasonal events such as Christmas and Valentine's Day.

[1113] The server then cleans and preprocesses the collected data. For example, it imputes missing values ​​in the sales data using the mean or median, and removes outliers. This preprocessing phase also scales the data and integrates sales data, market trend data, and seasonality data. The resulting dataset is then used to train machine learning models.

[1114] The server then uses the preprocessed data to train a machine learning model. Here, the data is split into training data and validation data, and a time series forecasting model or neural network model is selected and trained. For example, the server uses an LSTM (long short-term memory) model to train a model to predict sales demand for the next month. During the training process, the data is repeatedly fed into the model, and patterns are learned.

[1115] The system also incorporates an emotion engine that analyzes user emotions and optimizes marketing strategies based on those emotions. Specifically, the server uses customer data to analyze customer emotions in real time. For example, if a customer expresses positive emotions, the system may introduce new products or make cross-selling suggestions. Conversely, if a customer expresses negative emotions, the system may provide customer support or send discount coupons.

[1116] Once training is complete, the server uses the newly collected data to predict future sales trends and demand. These predictions are stored locally and then made available to users via a dashboard. For example, the dashboard might display sales forecasts for a specific product during the upcoming Christmas season.

[1117] Once the forecast results are available, the server then performs dynamic pricing, adjusting product prices in real time based on predicted demand. For example, discounts are applied to items with high inventory, and prices are increased for items predicted to be in high demand. This dynamic pricing allows businesses to maximize profits.

[1118] Furthermore, the device displays the forecast results to the user through a dashboard. Information is presented using visual elements (graphs, charts) for easy understanding. Users can also monitor the effectiveness of their marketing efforts in real time, allowing them to quickly determine the success of their efforts and any necessary adjustments.

[1119] As described above, the present invention provides a system that provides dynamic pricing based on highly accurate demand forecasts and an optimized marketing strategy using an emotion engine, contributing to corporate sales planning, inventory management, and profit maximization.

[1120] The processing flow will be explained below.

[1121] Step 1:

[1122] The server accesses the company's database and collects past sales data, such as monthly sales data for the past five years and detailed information such as the sales volume and price of each product.

[1123] Step 2:

[1124] The server uses external APIs to collect market trend data, specifically, Google Trends and social media data, to understand consumer interest trends in the market.

[1125] Step 3:

[1126] The server accesses a calendar event database and collects data that reflects seasonality, for example, collecting data about seasonal events such as Christmas and Valentine's Day.

[1127] Step 4:

[1128] The server cleans the collected data by detecting missing values ​​in the dataset, imputing them with the mean or median, and removing outliers.

[1129] Step 5:

[1130] The server pre-processes the cleaned data, including scaling and normalizing it, and integrating sales data, market trend data, and seasonality data.

[1131] Step 6:

[1132] The server splits the preprocessed dataset into training and validation data, typically 80% of the total data for training and 20% for validation.

[1133] Step 7:

[1134] The server selects a time series prediction model or neural network model, for example, a long short-term memory (LSTM) model.

[1135] Step 8:

[1136] The server uses the training data to train the machine learning model by repeatedly inputting the data into the model and letting it learn patterns.

[1137] Step 9:

[1138] The server evaluates the performance of the trained model using validation data, using metrics such as MSE (Mean Squared Error) and MAE (Mean Absolute Error).

[1139] Step 10:

[1140] The server uses the new data to predict future sales trends and demand, for example, predicting the sales volume of a particular product for the next month.

[1141] Step 11:

[1142] The server stores the prediction results in a database and generates a dashboard, which displays the prediction results in the form of graphs and charts.

[1143] Step 12:

[1144] The terminal displays the prediction results to the user through a dashboard, providing visual information that is easy for the user to understand.

[1145] Step 13:

[1146] The server dynamically sets prices based on the demand forecast results, for example, raising prices for products with high predicted demand and discounting products with high inventory.

[1147] Step 14:

[1148] The user implements the individually optimized marketing strategy provided by the server, specifically, sending personalized promotional emails to specific customer segments.

[1149] Step 15:

[1150] The server runs an emotion engine using customer data, such as past purchase history, browsing history, and social media comments, to analyze customer emotions in real time.

[1151] Step 16:

[1152] The server uses the emotion engine to generate more effective marketing strategies based on the obtained emotion data, such as introducing new products to customers who show positive emotions and providing support or discount coupons to customers who show negative emotions.

[1153] Step 17:

[1154] Users can monitor the effectiveness of their marketing initiatives in real time, allowing them to quickly determine the success of their initiatives and make any necessary adjustments.

[1155] Example 2

[1156] Next, a description will be given of 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."

[1157] For businesses, accurate demand forecasting and determining effective marketing strategies are important issues that directly affect sales growth and inventory management. However, many businesses are unable to properly utilize past data, making it difficult to accurately reflect seasonality and market trends. Formulating appropriate marketing strategies based on customer sentiment is also a challenge. The present invention aims to solve these problems and provide a set of tools that enable businesses to optimize sales planning, inventory management, and marketing strategies.

[1158] 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.

[1159] In this invention, the server includes means for collecting past transaction data, external market trend data, and seasonal information, means for cleaning and pre-processing the collected data, means for training a machine learning model for demand forecasting using the pre-processed data, means for forecasting future sales trends and demand using the trained model, means for generating a visualization interface for displaying the forecast results, means for dynamic pricing based on the forecast results, and means for analyzing customer sentiment data using a sentiment analyzer and generating individually optimized marketing strategies based on the results, thereby enabling highly accurate sales forecasting and optimization of marketing strategies based on customer sentiment.

[1160] "Past transaction data" refers to information about all sales activities conducted by a company in the past, including details such as sales quantity, sales price, and sales date and time.

[1161] "External market trend data" refers to information that shows overall market trends, competitor activities, consumer purchasing trends, etc., and is obtained from external information providers or APIs.

[1162] "Seasonal information" is data that indicates fluctuations in sales related to specific seasons or events, and includes sales trends during specific periods such as Christmas and Valentine's Day.

[1163] "Cleaning" refers to the process of removing errors, missing values, and inappropriate data from collected data to improve the quality of the data.

[1164] "Preprocessing" is the process of preparing data in a format suitable for a machine learning model, and includes scaling and standardizing the data, and filling in missing values.

[1165] A "machine learning model" is an algorithm or system used to make predictions, classifications, and optimizations based on input data, and includes LSTM and neural networks.

[1166] "Training" refers to the process of using existing data to teach a machine learning model so that it can make appropriate predictions and recommendations for new data.

[1167] "Future Sales Trends" refers to future sales trends and changes in demand predicted by the machine learning model.

[1168] "Prediction results" refer to the results of a machine learning model's predictions about the future, and include information such as specific sales volumes and demand patterns.

[1169] A "visualization interface" is a means for displaying prediction results in a visual format such as a graph or chart, making it easy for users to understand.

[1170] "Dynamic pricing" refers to the practice of adjusting product prices in real time based on predicted demand, raising or lowering prices based on inventory levels and supply-demand balances.

[1171] An "emotion analyzer" is a device or software that analyzes customer emotions and determines the emotional state of customers by analyzing reviews, social media posts, etc.

[1172] An "individually optimized marketing strategy" refers to a strategy that designs and executes optimal marketing methods for individual customers based on collected customer data and the results of sentiment analysis.

[1173] This invention is a system that uses AI and machine learning to predict future sales trends and demand, and further uses an emotion engine to generate marketing strategies based on user emotions. This system aims to help companies optimize their sales plans, inventory management, and marketing strategies.

[1174] Data collection

[1175] server

[1176] The server first accesses the company's database to retrieve transaction data from the past five years. This data includes details such as the sales quantity, sales price, and sales date for each product. It also collects market trend data through an external API. This market trend data includes consumer purchasing trends and competitor activity. The server also accesses a calendar event database to collect information about seasonal events such as Christmas and Valentine's Day.

[1177] Data Preprocessing

[1178] server

[1179] The server cleans and preprocesses the collected data. Specifically, if there are missing values ​​in the sales data, they are imputed using the mean or median. Outliers are also removed as soon as they are detected. The data is then scaled and converted into a unified format. Sales data, market trend data, and seasonal data are integrated into a single dataset.

[1180] Training a machine learning model

[1181] server

[1182] The preprocessed dataset is used to train the machine learning model. The dataset is divided into training data and validation data. The server selects an LSTM (long short-term memory) model and trains it to learn sales demand by inputting data into the model. This process is performed so that the model can learn sales patterns and improve its prediction accuracy for future data.

[1183] Predicting future sales trends

[1184] server

[1185] Once training is complete, the newly collected data is used to predict future sales trends and demand, such as predicting sales volumes for a particular product during the upcoming Christmas season. These predictions are stored on the server and later provided to users via a dashboard.

[1186] Utilizing the Emotion Engine

[1187] server

[1188] The server uses a sentiment analyzer to analyze user sentiment in real time. For example, it analyzes customer reviews and social media posts, and if the sentiment is positive, it introduces new products or suggests cross-selling. Conversely, if the sentiment is negative, it provides customer support or sends discount coupons.

[1189] Dynamic Pricing

[1190] server

[1191] Based on the predictions, the server dynamically sets prices, for example, applying discounts to items with high inventory and raising prices for items predicted to be in high demand, aiming to maximize the company's profits.

[1192] Viewing the Dashboard

[1193] Terminal

[1194] The dashboard displays the forecast results sent from the server, using visual elements such as graphs and charts to provide information that users can easily understand. Users can also monitor the effectiveness of their marketing efforts in real time, quickly determining the success of their efforts and any necessary adjustments.

[1195] Specific examples

[1196] For example, the server collects data for the Christmas shopping season in December. First, it collects sales data from past Christmas seasons from the database, then obtains market trend data using an external API. It also collects Christmas-related data from the calendar event database. The server then imputes any missing values ​​in the collected data using the average value, scales and integrates all the data. Once all the datasets are ready, it uses them to train an LSTM model. It then performs predictions on the newly acquired data and displays the results on a dashboard. Finally, it uses an emotion engine to introduce new products if customers have positive emotions.

[1197] Prompt Sentence Examples

[1198] "For the Christmas sales season leading up to December, collect Christmas season sales data, market trend data, and seasonal event data from the past five years, and use this data to make predictions using an LSTM model and display them on a dashboard. In addition, analyze customer sentiment data in real time, and if customers show positive sentiment, introduce new products or make cross-selling suggestions."

[1199] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1200] Step 1: Data collection

[1201] server

[1202] The server first accesses the company's database and retrieves transaction data from the past five years, including details such as the quantity, price, and date of each product sold.

[1203] Input: Company database

[1204] Output: Transaction data for the past 5 years

[1205] The server then collects market trend data through external APIs, including consumer purchasing habits and competitor activity.

[1206] Input: External API

[1207] Output: Market trend data

[1208] Additionally, the server accesses a calendar events database to collect information about seasonal events such as Christmas and Valentine's Day.

[1209] Input: Calendar event database

[1210] Output: Seasonal event data

[1211] Step 2: Data Preprocessing

[1212] server

[1213] The server cleans and preprocesses all collected data. Specifically, if there are missing values ​​in the sales data, they are imputed using the mean or median. Outliers are also removed as soon as they are detected. The server also scales the data and converts it into a unified format. Finally, the sales data, market price trend data, and seasonal data are integrated into a single dataset.

[1214] Inputs: trade data, market trend data, seasonal event data

[1215] Output: Preprocessed dataset

[1216] Step 3: Train the model

[1217] server

[1218] The preprocessed dataset is used to train a machine learning model. The dataset is divided into training data and validation data. The server selects an LSTM (long short-term memory) model and inputs the data multiple times to train the model on sales demand. This training process allows the model to learn sales patterns and improve its prediction accuracy for future data.

[1219] Input: Preprocessed dataset

[1220] Output: A trained LSTM model

[1221] Step 4: Predict future sales trends

[1222] server

[1223] Using the trained LSTM model, the server analyzes newly collected data and predicts future sales trends and demand. For example, it predicts the sales volume of a particular product during the upcoming Christmas season. The prediction results are stored on the server and later provided to users via a dashboard.

[1224] Input: Newly collected data, trained LSTM model

[1225] Output: Future sales trend forecast results

[1226] Step 5: Leverage the Emotion Engine

[1227] server

[1228] The server uses a sentiment analyzer to analyze user emotions in real time. Specifically, it analyzes customer reviews and social media posts, and introduces new products and makes cross-selling suggestions to customers who show positive emotions. Conversely, it provides customer support and sends discount coupons to customers who show negative emotions.

[1229] Input: Customer reviews, social media posts

[1230] Output: Emotion-based marketing strategies

[1231] Step 6: Dynamic Pricing

[1232] server

[1233] Based on the predictions, the server dynamically sets prices, applying discounts to items with high inventory and raising prices for items predicted to be in high demand, thereby maximizing profits for the company.

[1234] Input: Future sales trend forecast results

[1235] Output: Real-time adjusted product prices

[1236] Step 7: View the dashboard

[1237] Terminal

[1238] The dashboard displays the prediction results sent from the server. The dashboard uses visual elements (graphs, charts) to provide users with easy-to-understand information. Users can also monitor the effectiveness of their marketing initiatives in real time through the dashboard, allowing them to quickly determine the success of their initiatives and any necessary adjustments.

[1239] Input: Future sales trend prediction results, emotion-based marketing strategies

[1240] Output: Visually displayed forecast results and strategies

[1241] (Application example 2)

[1242] Next, a description will be given of Application 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."

[1243] Modern marketing strategies and sales plans must respond quickly and accurately to diverse customer needs. However, traditional systems are based solely on past sales data and market trend data, which means they cannot fully consider customer sentiment or real-time conditions. Other issues include the difficulty of improving the accuracy of demand forecasts and dynamic pricing, and providing individually optimized marketing strategies in real time. This makes it difficult for companies to optimize sales plans and inventory management, leading to a demand for systems that contribute to maximizing profits.

[1244] 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.

[1245] In this invention, the server includes means for collecting past sales data, external market trend data, and seasonal data, means for cleaning and pre-processing the collected data, means for training a machine learning model for demand forecasting using the pre-processed data, means for collecting and analyzing customer sentiment data, and means for optimizing marketing strategies based on customer sentiment, thereby improving the accuracy of demand forecasting and marketing strategies and enabling optimal sales planning and inventory management that take customer sentiment into account.

[1246] "Sales data" refers to historical data relating to past sales, inventory information, product sales quantities and prices, and the like.

[1247] "Market trend data" refers to data on customer purchasing trends, popular product trends, and economic conditions in a broad range of markets.

[1248] "Seasonal data" refers to data on fluctuations in customer purchasing behavior associated with specific seasons or events, such as Christmas or Valentine's Day.

[1249] A "machine learning model" is a collection of algorithms that learn patterns from data and make predictions or classifications based on those patterns.

[1250] "Emotional data" refers to data on the emotional movements and feelings of customers that can be obtained from their facial expressions, voice, and behavior.

[1251] A "marketing strategy" is a plan for a series of advertising, promotion, pricing, and other activities designed to achieve a specific objective.

[1252] "Training data" is a learning dataset used when building a machine learning model.

[1253] "Validation data" is a dataset used to evaluate the performance of a trained machine learning model.

[1254] A "dashboard" is an interface that visually displays forecast results and other important information in a format that is easy for users to understand.

[1255] "Dynamic pricing" is a method of adjusting product prices in real time based on demand forecasts and inventory status.

[1256] "Demand forecasting" refers to predicting future customer purchasing behavior and sales volumes based on past data and trend information.

[1257] As an embodiment of the present invention, a server uses AI and machine learning to predict future sales trends and demand, and also uses an emotion engine to generate a marketing strategy based on user emotions. Specific implementation means are described below.

[1258] Data collection

[1259] The server accesses the company's database to collect past sales data, external market trend data, and seasonal data, including the "sales data," "market trend data," and "seasonal data" previously specified. For example, the server uses an API to obtain external trend data and retrieves past sales history from the company's internal database.

[1260] Data Cleaning and Preprocessing

[1261] The collected data is cleaned, missing values ​​are filled, and outliers are removed. This is a necessary step to increase the reliability of the data. For example, missing values ​​are filled with the mean or median. The data is then scaled and sales data, market trend data, and seasonal data are integrated.

[1262] Training a machine learning model

[1263] The preprocessed data is used to train a machine learning model. The model used here is a "long short-term memory (LSTM) model" or a "neural network model." The server splits the data into training data and validation data and performs training.

[1264] Demand forecasting

[1265] The trained model is used to predict future sales trends and demand, for example, predicting demand for specific products around seasonal events like Christmas and Valentine's Day. These predictions are then displayed on a dashboard as graphs and charts that users can easily understand.

[1266] Emotion data collection and analysis

[1267] The server collects customer emotional data from the smartphone's camera and microphone. Using specialized software for emotion analysis, it analyzes facial expressions and voice in real time. This emotional data is used to optimize marketing strategies.

[1268] Optimizing your marketing strategy

[1269] Optimize your marketing strategies based on customer sentiment data. For example, if a customer expresses positive sentiment, introduce new products or offer cross-selling suggestions. If a customer expresses negative sentiment, offer customer support or send discount coupons.

[1270] Dynamic Pricing

[1271] Dynamically adjust pricing based on demand forecasts. If demand for a particular product is predicted to increase, prices are raised, while discounts are offered if inventory is high. This allows for real-time price adjustments to maximize profits.

[1272] Dashboard View

[1273] It generates a dashboard that visually displays forecast results, sentiment analysis results, and pricing information, allowing users to easily check upcoming sales plans and marketing initiatives and quickly adjust them as needed.

[1274] For example, the following prompt sentence is input to the generative AI model:

[1275] "Please forecast the demand for home video game consoles during the next Christmas season."

[1276] "Analyze recent user reviews to identify products that have received positive feedback."

[1277] This allows companies to develop highly accurate sales strategies based on market trends and customer sentiment, and implement optimized marketing strategies.

[1278] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1279] Step 1:

[1280] The server collects historical sales data from the company's database and obtains external market trend data and seasonal data through API. The input is database connection information and API key, and the output is the integrated raw data set. Specifically, it issues database queries and API requests and stores the obtained data in local storage.

[1281] Step 2:

[1282] The server performs data cleaning on the collected raw dataset. The input is the integrated raw dataset, and the output is a clean dataset. Specific operations include imputing missing values ​​(calculating the mean and median), removing outliers, and performing data scaling. This results in a dataset suitable for training machine learning models.

[1283] Step 3:

[1284] The server trains a machine learning model using a clean dataset. The input is the clean dataset, and the output is the trained model. Specifically, the data is split into training data and validation data, and training is performed using an LSTM model or a neural network model. Machine learning libraries such as TensorFlow and Keras are used to train the model.

[1285] Step 4:

[1286] The server uses the trained model to predict future sales trends and demand. The input is a newly collected dataset, and the output is the prediction result. Specifically, new data is input into the model, and predictions are made to calculate demand fluctuations and sales trends.

[1287] Step 5:

[1288] The device collects user emotional data through the smartphone's camera and microphone. The input is video and audio data, and the output is the result of emotional analysis. Specifically, it uses an emotion analysis engine to analyze changes in facial expressions and voice tone in real time to extract emotional data.

[1289] Step 6:

[1290] The server generates an individually optimized marketing strategy based on the emotional data. The input is the result of the emotional analysis, and the output is a set of marketing measures. Specifically, it runs an algorithm that performs promotions and product recommendations according to the emotional state.

[1291] Step 7:

[1292] The server performs dynamic pricing based on the demand forecast results. The inputs are the demand forecast results and inventory data, and the output is adjusted price information. Specifically, it runs an algorithm that raises prices when demand is high and applies discounts when inventory is high.

[1293] Step 8:

[1294] The terminal generates a dashboard and displays the forecast results, sentiment analysis results, and pricing information. The input is the forecast results and sentiment analysis results, and the output is a visually displayed dashboard. Specific operations include displaying information using graphs and charts, allowing users to easily check sales forecasts and marketing strategies.

[1295] Through these steps, the server and terminals work together to provide a system that realizes highly accurate demand forecasting, dynamic pricing, and emotion-based marketing strategies.

[1296] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 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.

[1297] 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> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) 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.

[1298] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1299] [Fourth embodiment]

[1300] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1301] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1302] 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).

[1303] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. 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 control target 443 are also connected to the bus 52.

[1304] 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.

[1305] 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).

[1306] 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.

[1307] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1308] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, 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.

[1309] 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.

[1310] 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.

[1311] In the robot 414, the processor 46 performs the reception output process. 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.

[1312] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1313] As an embodiment of the present invention, we provide a system that uses AI and machine learning to forecast future sales trends and demand. This system aims to enable companies to effectively optimize sales planning, inventory management, and marketing strategies by using past sales data, market trend data, and seasonal data.

[1314] The specific program processing of this system is as follows.

[1315] First, the server accesses the company's database to collect past sales data, external market trend data, and seasonal data. For example, the server obtains monthly sales data for the past five years and then obtains market trend data from an external API. In this way, multifaceted data collection forms the basis for a more accurate forecasting model.

[1316] The server then cleans and preprocesses the collected data. For example, missing values ​​in the sales data are imputed using the mean or median. Outliers are also removed. This preprocessing phase normalizes the data and integrates market trend and seasonal data with the sales data. The resulting dataset is then used to train the machine learning model.

[1317] The server then uses the preprocessed data to train a machine learning model. Here, the data is split into training data and validation data, and a time series forecasting model or neural network model is selected and trained. For example, the server uses an LSTM (long short-term memory) model to train a model to predict sales demand for the next month. During this training process, the data is repeatedly fed into the model, and patterns are learned.

[1318] Once training is complete, the server uses the newly collected data to predict future sales trends and demand. These predictions are stored locally and then made available to users via a dashboard. For example, the dashboard might display sales forecasts for a specific product during the upcoming Christmas season.

[1319] Once the forecast results are available, the server then performs dynamic pricing, adjusting product prices in real time based on predicted demand. For example, discounts are applied to items with high inventory, and prices are increased for items predicted to be in high demand. This dynamic pricing allows businesses to maximize profits.

[1320] Furthermore, the user implements individually optimized marketing strategies provided by the server. The server generates personalized promotional emails for specific customer segments based on the prediction results and customer data. The user can then run advertising campaigns based on this and monitor the results in real time.

[1321] In this way, the present invention provides dynamic pricing and optimized marketing strategies based on highly accurate demand forecasts, realizing a system that contributes to corporate sales planning, inventory management, and profit maximization.

[1322] The processing flow will be explained below.

[1323] Step 1:

[1324] The server accesses the company's database and collects past sales data, such as monthly sales data for the past five years and detailed information such as the sales volume and price of each product.

[1325] Step 2:

[1326] The server uses external APIs to collect relevant market trend data, specifically Google Trends and social media data, to understand consumer interest trends in the market.

[1327] Step 3:

[1328] The server accesses a calendar event database to retrieve data that reflects seasonality, for example, collecting data about seasonal events such as Christmas and Valentine's Day.

[1329] Step 4:

[1330] The server cleans the collected data by detecting missing values ​​in the dataset, imputing them with the mean or median, and removing outliers.

[1331] Step 5:

[1332] The server pre-processes the cleaned data, including scaling and normalizing the data, and integrating sales data, market trend data, and seasonality data.

[1333] Step 6:

[1334] The server splits the preprocessed dataset into training and validation data, typically 80% of the total data for training and 20% for validation.

[1335] Step 7:

[1336] The server selects a time series prediction model or neural network model, for example, a long short-term memory (LSTM) model.

[1337] Step 8:

[1338] The server uses the training data to train the machine learning model, which involves repeatedly feeding the data into the model and letting it learn patterns.

[1339] Step 9:

[1340] The server evaluates the performance of the trained model using validation data, using metrics such as MSE (Mean Squared Error) and MAE (Mean Absolute Error).

[1341] Step 10:

[1342] The server uses the new data to predict future sales trends and demand, for example, predicting the sales volume of a particular product for the next month.

[1343] Step 11:

[1344] The server stores the prediction results in a database and generates a dashboard, which displays the prediction results in the form of graphs and charts.

[1345] Step 12:

[1346] The device displays the prediction results to the user through a dashboard, providing visual information that is easy for the user to understand.

[1347] Step 13:

[1348] The server dynamically sets prices based on the demand forecast results, for example, raising prices for products with high predicted demand and discounting products with high inventory.

[1349] Step 14:

[1350] The user implements individually optimized marketing strategies generated by the server, for example, sending personalized promotional emails to specific customer segments.

[1351] Step 15:

[1352] Users can monitor the effectiveness of their marketing efforts in real time, allowing them to quickly determine the success of their efforts and make any necessary adjustments.

[1353] Example 1

[1354] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1355] Conventional sales forecasting systems often base their forecasts solely on past sales data and are unable to comprehensively incorporate external market trend data or seasonal data, resulting in low forecast accuracy. Furthermore, they lack effective means for implementing dynamic pricing and individually optimized marketing strategies. A system that can solve this issue and provide highly accurate demand forecasts and dynamic pricing and marketing strategies based on those forecasts is needed.

[1356] 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.

[1357] In this invention, the server includes means for collecting past sales data, external market trend data, and seasonal data, means for cleaning and pre-processing the collected data, means for training a machine learning model for demand forecasting using the pre-processed data, means for forecasting future sales trends and demand using the trained model, means for generating a recording medium for displaying the forecast results, means for dynamic pricing based on the forecast results, and means for generating an individually optimized marketing strategy based on customer data, thereby enabling highly accurate demand forecasting, dynamic pricing, and the implementation of individually optimized marketing strategies.

[1358] "Sales data" is a collection of information about sales performance of a particular product over a past period.

[1359] "Market trend data" is a collection of information that indicates the current situation and future trends in a particular market or industry.

[1360] "Seasonal data" is a collection of information that indicates sales trends and demand fluctuations associated with particular seasons or annual events.

[1361] "Data cleaning" is the process of removing inaccurate, incomplete, or unnecessary values ​​from collected data to improve data quality.

[1362] "Preprocessing" refers to a series of steps taken to convert collected data into a format suitable for analysis and machine learning.

[1363] A "machine learning model" is an algorithm or mathematical model that learns patterns from data and performs tasks such as prediction and classification.

[1364] "Training" is the process of providing a machine learning model with large amounts of data and having it learn from that data to improve its prediction accuracy.

[1365] A "recording medium" is a physical or electronic device or means for storing and displaying digital data.

[1366] "Dynamic pricing" is a method of adjusting prices in real time based on demand forecasts and market conditions.

[1367] A "marketing strategy" is a series of plans and measures implemented to promote sales and increase brand awareness among specific customer segments.

[1368] This invention relates to a sales forecasting system, specifically a system that uses AI and machine learning technologies to forecast future sales trends and demand. This system is designed to help companies more effectively plan and execute sales plans, inventory management, and marketing strategies by utilizing past sales data, market trend data, and seasonal data.

[1369] Hardware and software used

[1370] 1. Server: Manages all data collection, preprocessing, machine learning model training, and dynamic pricing using libraries such as Python, Pandas, Scikit-learn, TensorFlow, or PyTorch.

[1371] 2. Database: Stores and manages company sales data and external trend data. Specifically, MySQL or PostgreSQL can be used.

[1372] 3. External API: Used to obtain market trend data.

[1373] 4. Web dashboard: An interface for providing predictions and analysis results to users. Specifically, JavaScript, D3.js, or Tableau may be used.

[1374] Data collection and preprocessing

[1375] The server accesses the company's database to collect past sales data, external market trend data, and seasonal data. This data allows for more accurate forecasts based on multifaceted information. Specifically, the server periodically sends API requests to collect and store the necessary data.

[1376] Next, the server cleans and preprocesses the data. This step includes imputing missing values, removing outliers, and normalizing the data. For example, it uses the Pandas fillna() method to impute missing values ​​with the mean.

[1377] Training a machine learning model

[1378] The server uses the preprocessed data to train a machine learning model. Here, the data is divided into training data and validation data, and a long short-term memory (LSTM) model is used to perform time series prediction. Specifically, an LSTM model is built and trained using TensorFlow.

[1379] Future sales trends and demand forecasts

[1380] Using the trained model, the server predicts future sales trends and demand based on newly collected data. These predictions are stored in a database and made available to users through a web dashboard. For example, the dashboard might predict that demand for a particular product will increase during the upcoming Christmas season.

[1381] Dynamic Pricing

[1382] The server then dynamically sets prices based on the forecast, for example, discounting items with high inventory and raising prices for items predicted to be in high demand. This dynamic pricing is a key factor in helping companies maximize profits.

[1383] Individually optimized marketing strategies

[1384] Based on the prediction results and customer data provided by the server, users can create personalized promotional emails and advertising campaigns for specific customer segments. For example, emails can be automatically generated to strongly appeal to specific customer demographics about a specific product.

[1385] Prompt Sentence Examples

[1386] "Please provide us with your sales forecast for electronic products for the upcoming Christmas season. Also, please identify which products will sell best and suggest a corresponding marketing strategy."

[1387] In this way, this system provides dynamic pricing and optimized marketing strategies based on highly accurate demand forecasts, contributing to corporate sales planning, inventory management, and profit maximization.

[1388] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1389] Step 1:

[1390] Data collection

[1391] The server accesses the company's database to collect historical sales data, external market trend data, and seasonality data.

[1392] Specific behavior:

[1393] The server periodically executes SQL queries to extract sales data from the company's database. The server sends HTTP requests to retrieve market trend data from an external API. The server stores this data in a local database.

[1394] input:

[1395] Corporate database connection information, external API endpoints.

[1396] output:

[1397] Collection of historical sales data, market trend data, and seasonality data.

[1398] Step 2:

[1399] Data Cleaning and Preprocessing

[1400] The server cleans the collected data, filling in missing values, removing outliers, and normalizing it.

[1401] Specific behavior:

[1402] The server uses the Pandas library to load the data into a data frame, uses the fillna() method to impute missing values, boxplots and Z-scores to detect and remove outliers, and normalizes each data set to create a unified dataset.

[1403] input:

[1404] Collected sales data, market trend data, and seasonality data.

[1405] output:

[1406] Cleaned and preprocessed integrated dataset.

[1407] Step 3:

[1408] Data partitioning

[1409] The server splits the combined dataset into training data and validation data.

[1410] Specific behavior:

[1411] The server splits the data into training and validation data using Scikit-learn's train_test_split function.

[1412] input:

[1413] Cleaned and preprocessed integrated dataset.

[1414] output:

[1415] Training and validation data.

[1416] Step 4:

[1417] Training a machine learning model

[1418] The server trains the LSTM model using the training data.

[1419] Specific behavior:

[1420] The server uses the TensorFlow library to build an LSTM model and trains it using the model.fit() method. During the training process, the validation data is used to evaluate the accuracy of the model.

[1421] input:

[1422] Training data, validation data.

[1423] output:

[1424] A trained LSTM model.

[1425] Step 5:

[1426] Future predictions

[1427] The server uses the trained model to predict future sales trends and demand.

[1428] Specific behavior:

[1429] The server preprocesses the newly collected data and inputs it into the trained LSTM model, using the model.predict() method to predict the sales demand for the next month.

[1430] input:

[1431] Newly collected data, trained LSTM model.

[1432] output:

[1433] Sales trends and demand forecast results.

[1434] Step 6:

[1435] Saving and displaying prediction results

[1436] The server stores the prediction results in a database and displays them on a dashboard that users can access.

[1437] Specific behavior:

[1438] The server stores the prediction results in a database and provides them to users via a web dashboard, where users can check the prediction results.

[1439] input:

[1440] Sales trends and demand forecast results.

[1441] output:

[1442] Prediction results displayed on a dashboard.

[1443] Step 7:

[1444] Dynamic Pricing

[1445] The server dynamically sets prices based on the prediction results.

[1446] Specific behavior:

[1447] The server runs a price optimization algorithm and adjusts prices based on inventory and demand forecasts, then feeds the adjusted prices into the company's ERP system.

[1448] input:

[1449] Sales trends and demand forecast results, inventory data.

[1450] output:

[1451] Dynamically adjusted pricing information.

[1452] Step 8:

[1453] Implementing marketing strategies

[1454] The user executes an individually optimized marketing campaign based on the marketing strategy provided by the server.

[1455] Specific behavior:

[1456] The server automatically generates promotional emails and advertising campaigns and sends them to specific customer segments, and users can monitor the effectiveness of the campaigns in real time using a dashboard.

[1457] input:

[1458] Prediction results, customer data.

[1459] output:

[1460] Individually optimized marketing strategies and execution results.

[1461] These are the specific processing steps of the program for this system, which enables companies to achieve highly accurate demand forecasting, dynamic pricing, and individually optimized marketing strategies.

[1462] (Application example 1)

[1463] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1464] Traditional sales trend and demand forecasting systems often lack sufficient forecasting accuracy, making them ineffective for companies' inventory management and marketing strategies. Furthermore, these systems require large desktop environments and servers and lack real-time responsiveness. As a result, it is difficult for brick-and-mortar store operators to respond quickly to changing market conditions and special seasonal events. Therefore, there is a need for a system that can achieve more accurate forecasting and dynamic pricing, thereby improving the operational efficiency of brick-and-mortar stores.

[1465] 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.

[1466] In this invention, the server includes means for collecting past sales data, external market trend data, and seasonal data, means for cleaning and preprocessing the collected data, means for training a machine learning model for demand forecasting using the preprocessed data, means for forecasting future sales trends and demand using the trained model, means for generating a dashboard for displaying the forecast results, means for dynamic pricing based on the forecast results, means for generating individually optimized marketing strategies based on customer data, and means for providing a smartphone application that performs real-time demand forecasting and dynamic pricing. This enables physical store operators to perform highly accurate demand forecasting and rapid price adaptation, thereby optimizing inventory management and marketing strategies.

[1467] "Past sales data" refers to numerical data and records relating to sales that have occurred in the past.

[1468] "External market trend data" refers to data obtained from external sources that indicates market trends and tendencies.

[1469] "Seasonal data" refers to data that indicates seasonal fluctuations and includes information such as consumer behavior at specific times.

[1470] "Means of collection" refers to the methods, tools, and processes used to obtain the necessary data.

[1471] "Cleaning" is a data preparation process that involves tasks such as deleting unnecessary elements and filling in missing values.

[1472] "Preprocessing means" refers to methods and tools used to prepare data in a format suitable for machine learning models.

[1473] "Demand forecasting" refers to estimating the future number of products that will be needed and providing information that can be incorporated into sales plans.

[1474] "Means for training a machine learning model" means the method or process of running a machine learning algorithm on collected data to build the model needed to make a prediction.

[1475] A "trained model" refers to a predictive model that has been learned by a machine learning algorithm using historical data.

[1476] "Means for generating dashboards" refers to methods for creating interfaces and tools that allow users to visually examine data.

[1477] "Dynamic pricing" is the process of adjusting prices in real time based on demand and other factors.

[1478] "Customer data" refers to information about individual customers, including purchase history and behavioral data.

[1479] An "individually optimized marketing strategy" refers to a sales promotion strategy that is customized to effectively reach a specific customer segment.

[1480] "Real-time demand forecasting" refers to systems and tools that can process data in real time and forecast demand instantly.

[1481] A "smartphone application" refers to software that runs on a mobile phone terminal and provides specific functions and services to users.

[1482] This invention provides a system that enables brick-and-mortar store operators to monitor sales trends and demand in real time and implement dynamic pricing and personalized marketing strategies. Specific embodiments for implementing this system are described below.

[1483] First, the server accesses the company's database to collect past sales data, external market trend data, and seasonal data, thereby providing the basis for highly accurate demand forecasting.

[1484] This collected data is cleaned and pre-processed on the server. Specifically, missing values ​​in the sales data are imputed using the mean and median, and outliers are removed. At this stage, the data is standardized and integrated with market trend and seasonal data.

[1485] The preprocessed data is used to train a machine learning model for demand forecasting. The server divides the data into training data and validation data, selects a time series forecasting model or neural network model, and begins training. For example, by using an LSTM (long short-term memory) model, a model that can accurately predict the next month's sales demand is built.

[1486] Furthermore, the server uses the newly collected data to forecast future sales trends and demand, and provides these forecast results to users via a dashboard, an interface that allows store operators to visually view the demand forecast results.

[1487] Users can then use these forecasts to implement dynamic pricing, which adjusts product prices in real time based on predicted demand. For example, they can maximize profits by applying discounts to items with high inventory and raising prices for items predicted to be in high demand.

[1488] Furthermore, the server automatically generates personalized marketing strategies based on the prediction results and customer data, which are tailored to specific customer segments and can, for example, send promotional emails for specific products to customers.

[1489] As part of the system, a smartphone application that performs real-time demand forecasting and dynamic pricing will also be provided, allowing store operators to easily check demand forecast results and adjust pricing on the go.

[1490] The hardware and software used are as follows: For hardware, a standard server machine and a smartphone terminal are used. For software, Python, Pandas, NumPy, Scikit-learn, Keras, etc. are used. Using these tools, data collection, cleaning, preprocessing, model training, and prediction can be performed efficiently.

[1491] A concrete example is a scenario in which the system needs to predict the demand for each product this Christmas season based on sales figures from last December, and then make pricing suggestions based on that. The user can request this by entering the following prompt:

[1492] "Based on last December's sales figures, please predict the demand for each product during this year's Christmas season. Also, please make pricing suggestions based on your results."

[1493] In this way, the present invention not only dramatically improves the operational efficiency of physical stores, but also realizes highly accurate demand forecasting and dynamic price adjustments.

[1494] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1495] Step 1:

[1496] A user launches a smartphone application and submits a request to collect historical sales data, external market trend data, and seasonality data.

[1497] Input: Request data

[1498] Output: Collection request

[1499] Specific Operation: When a user presses a button on the application to start data collection, a request is sent to the server to collect historical sales data, external market trend data, and seasonal data.

[1500] Step 2:

[1501] The server accesses the company's database to retrieve historical sales data, collects market trend data from external APIs, and loads pre-defined seasonality data.

[1502] Input: Collection request

[1503] Output: Raw data

[1504] What happens: The server queries the company's database to retrieve historical sales data, and simultaneously sends a request to an external API to retrieve market trend data and load seasonal data files.

[1505] Step 3:

[1506] The server cleans and pre-processes the collected data, imputing missing values, removing outliers, and standardizing the data. It also integrates market trend and seasonal data with sales data.

[1507] Input: Raw data

[1508] Output: Preprocessed data

[1509] What it does: It uses Pandas to impute missing data values ​​with the mean or median, detect and remove outliers, and standardize and combine data into a single dataset.

[1510] Step 4:

[1511] The server uses the preprocessed data to train a machine learning model for demand forecasting. It splits the data into training data and validation data and uses a time series model such as LSTM.

[1512] Input: Preprocessed data

[1513] Output: A trained model

[1514] Specific operation: Using Scikit-learn and Keras, the data is divided into training data and validation data, and an LSTM model is constructed and trained.

[1515] Step 5:

[1516] The server uses the newly collected data to predict future sales trends and demand with a trained model.

[1517] Input: New raw data

[1518] Output: Demand forecast results

[1519] Specific operation: After preprocessing the newly collected sales data and market trend data, it is input into the trained model to predict future demand.

[1520] Step 6:

[1521] The server dynamically sets prices using a pricing algorithm based on the prediction results.

[1522] Input: Demand forecast results

[1523] Output: Pricing data

[1524] Specific operation: Based on the prediction results, the server performs dynamic pricing, lowering the prices of items with high inventory and raising the prices of items with high demand.

[1525] Step 7:

[1526] The forecast results and pricing data are generated as a dashboard and displayed on a smartphone application.

[1527] Input: Forecast results, pricing data

[1528] Output: Dashboard

[1529] Specific operation: The server visualizes the forecast results and pricing data, generates a dashboard, and delivers it to a smartphone application.

[1530] Step 8:

[1531] Users can check the dashboard through a smartphone application and adjust prices and manage inventory in physical stores based on sales strategies.

[1532] Input: Dashboard

[1533] Output: Sales strategy implementation

[1534] Specific Action: User reviews the application dashboard and implements specific sales strategies based on predicted demand and pricing.

[1535] Through this process, physical store operators can achieve highly accurate demand forecasts and quick price adjustments.

[1536] 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.

[1537] As an embodiment of the present invention, we provide a system that uses AI and machine learning to predict future sales trends and demand, and generates marketing strategies based on user emotions using an emotion engine. This system utilizes past sales data, market trend data, seasonal data, and customer emotion data to enable companies to effectively optimize sales plans, inventory management, and marketing strategies.

[1538] First, the server accesses the company's database to collect past sales data, external market trend data, and seasonal data. For example, it obtains detailed information such as monthly sales data for the past five years and the sales volume and price of each product. It also obtains market trend data from an external API. It also accesses a calendar event database to collect data on seasonal events such as Christmas and Valentine's Day.

[1539] The server then cleans and preprocesses the collected data. For example, it imputes missing values ​​in the sales data using the mean or median, and removes outliers. This preprocessing phase also scales the data and integrates sales data, market trend data, and seasonality data. The resulting dataset is then used to train machine learning models.

[1540] The server then uses the preprocessed data to train a machine learning model. Here, the data is split into training data and validation data, and a time series forecasting model or neural network model is selected and trained. For example, the server uses an LSTM (long short-term memory) model to train a model to predict sales demand for the next month. During the training process, the data is repeatedly fed into the model, and patterns are learned.

[1541] The system also incorporates an emotion engine that analyzes user emotions and optimizes marketing strategies based on those emotions. Specifically, the server uses customer data to analyze customer emotions in real time. For example, if a customer expresses positive emotions, the system may introduce new products or make cross-selling suggestions. Conversely, if a customer expresses negative emotions, the system may provide customer support or send discount coupons.

[1542] Once training is complete, the server uses the newly collected data to predict future sales trends and demand. These predictions are stored locally and then made available to users via a dashboard. For example, the dashboard might display sales forecasts for a specific product during the upcoming Christmas season.

[1543] Once the forecast results are available, the server then performs dynamic pricing, adjusting product prices in real time based on predicted demand. For example, discounts are applied to items with high inventory, and prices are increased for items predicted to be in high demand. This dynamic pricing allows businesses to maximize profits.

[1544] Furthermore, the device displays the forecast results to the user through a dashboard. Information is presented using visual elements (graphs, charts) for easy understanding. Users can also monitor the effectiveness of their marketing efforts in real time, allowing them to quickly determine the success of their efforts and any necessary adjustments.

[1545] As described above, the present invention provides a system that provides dynamic pricing based on highly accurate demand forecasts and an optimized marketing strategy using an emotion engine, contributing to corporate sales planning, inventory management, and profit maximization.

[1546] The processing flow will be explained below.

[1547] Step 1:

[1548] The server accesses the company's database and collects past sales data, such as monthly sales data for the past five years and detailed information such as the sales volume and price of each product.

[1549] Step 2:

[1550] The server uses external APIs to collect market trend data, specifically, Google Trends and social media data, to understand consumer interest trends in the market.

[1551] Step 3:

[1552] The server accesses a calendar event database and collects data that reflects seasonality, for example, collecting data about seasonal events such as Christmas and Valentine's Day.

[1553] Step 4:

[1554] The server cleans the collected data by detecting missing values ​​in the dataset, imputing them with the mean or median, and removing outliers.

[1555] Step 5:

[1556] The server pre-processes the cleaned data, including scaling and normalizing it, and integrating sales data, market trend data, and seasonality data.

[1557] Step 6:

[1558] The server splits the preprocessed dataset into training and validation data, typically 80% of the total data for training and 20% for validation.

[1559] Step 7:

[1560] The server selects a time series prediction model or neural network model, for example, a long short-term memory (LSTM) model.

[1561] Step 8:

[1562] The server uses the training data to train the machine learning model by repeatedly inputting the data into the model and letting it learn patterns.

[1563] Step 9:

[1564] The server evaluates the performance of the trained model using validation data, using metrics such as MSE (Mean Squared Error) and MAE (Mean Absolute Error).

[1565] Step 10:

[1566] The server uses the new data to predict future sales trends and demand, for example, predicting the sales volume of a particular product for the next month.

[1567] Step 11:

[1568] The server stores the prediction results in a database and generates a dashboard, which displays the prediction results in the form of graphs and charts.

[1569] Step 12:

[1570] The terminal displays the prediction results to the user through a dashboard, providing visual information that is easy for the user to understand.

[1571] Step 13:

[1572] The server dynamically sets prices based on the demand forecast results, for example, raising prices for products with high predicted demand and discounting products with high inventory.

[1573] Step 14:

[1574] The user implements the individually optimized marketing strategy provided by the server, specifically, sending personalized promotional emails to specific customer segments.

[1575] Step 15:

[1576] The server runs an emotion engine using customer data, such as past purchase history, browsing history, and social media comments, to analyze customer emotions in real time.

[1577] Step 16:

[1578] The server uses the emotion engine to generate more effective marketing strategies based on the obtained emotion data, such as introducing new products to customers who show positive emotions and providing support or discount coupons to customers who show negative emotions.

[1579] Step 17:

[1580] Users can monitor the effectiveness of their marketing initiatives in real time, allowing them to quickly determine the success of their initiatives and make any necessary adjustments.

[1581] Example 2

[1582] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1583] For businesses, accurate demand forecasting and determining effective marketing strategies are important issues that directly affect sales growth and inventory management. However, many businesses are unable to properly utilize past data, making it difficult to accurately reflect seasonality and market trends. Formulating appropriate marketing strategies based on customer sentiment is also a challenge. The present invention aims to solve these problems and provide a set of tools that enable businesses to optimize sales planning, inventory management, and marketing strategies.

[1584] 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.

[1585] In this invention, the server includes means for collecting past transaction data, external market trend data, and seasonal information, means for cleaning and pre-processing the collected data, means for training a machine learning model for demand forecasting using the pre-processed data, means for forecasting future sales trends and demand using the trained model, means for generating a visualization interface for displaying the forecast results, means for dynamic pricing based on the forecast results, and means for analyzing customer sentiment data using a sentiment analyzer and generating individually optimized marketing strategies based on the results, thereby enabling highly accurate sales forecasting and optimization of marketing strategies based on customer sentiment.

[1586] "Past transaction data" refers to information about all sales activities conducted by a company in the past, including details such as sales quantity, sales price, and sales date and time.

[1587] "External market trend data" refers to information that shows overall market trends, competitor activities, consumer purchasing trends, etc., and is obtained from external information providers or APIs.

[1588] "Seasonal information" is data that indicates fluctuations in sales related to specific seasons or events, and includes sales trends during specific periods such as Christmas and Valentine's Day.

[1589] "Cleaning" refers to the process of removing errors, missing values, and inappropriate data from collected data to improve the quality of the data.

[1590] "Preprocessing" is the process of preparing data in a format suitable for a machine learning model, and includes scaling and standardizing the data, and filling in missing values.

[1591] A "machine learning model" is an algorithm or system used to make predictions, classifications, and optimizations based on input data, and includes LSTM and neural networks.

[1592] "Training" refers to the process of using existing data to teach a machine learning model so that it can make appropriate predictions and recommendations for new data.

[1593] "Future Sales Trends" refers to future sales trends and changes in demand predicted by the machine learning model.

[1594] "Prediction results" refer to the results of a machine learning model's predictions about the future, and include information such as specific sales volumes and demand patterns.

[1595] A "visualization interface" is a means for displaying prediction results in a visual format such as a graph or chart, making it easy for users to understand.

[1596] "Dynamic pricing" refers to the practice of adjusting product prices in real time based on predicted demand, raising or lowering prices based on inventory levels and supply-demand balances.

[1597] An "emotion analyzer" is a device or software that analyzes customer emotions and determines the emotional state of customers by analyzing reviews, social media posts, etc.

[1598] An "individually optimized marketing strategy" refers to a strategy that designs and executes optimal marketing methods for individual customers based on collected customer data and the results of sentiment analysis.

[1599] This invention is a system that uses AI and machine learning to predict future sales trends and demand, and further uses an emotion engine to generate marketing strategies based on user emotions. This system aims to help companies optimize their sales plans, inventory management, and marketing strategies.

[1600] Data collection

[1601] server

[1602] The server first accesses the company's database to retrieve transaction data from the past five years. This data includes details such as the sales quantity, sales price, and sales date for each product. It also collects market trend data through an external API. This market trend data includes consumer purchasing trends and competitor activity. The server also accesses a calendar event database to collect information about seasonal events such as Christmas and Valentine's Day.

[1603] Data Preprocessing

[1604] server

[1605] The server cleans and preprocesses the collected data. Specifically, if there are missing values ​​in the sales data, they are imputed using the mean or median. Outliers are also removed as soon as they are detected. The data is then scaled and converted into a unified format. Sales data, market trend data, and seasonal data are integrated into a single dataset.

[1606] Training a machine learning model

[1607] server

[1608] The preprocessed dataset is used to train the machine learning model. The dataset is divided into training data and validation data. The server selects an LSTM (long short-term memory) model and trains it to learn sales demand by inputting data into the model. This process is performed so that the model can learn sales patterns and improve its prediction accuracy for future data.

[1609] Predicting future sales trends

[1610] server

[1611] Once training is complete, the newly collected data is used to predict future sales trends and demand, such as predicting sales volumes for a particular product during the upcoming Christmas season. These predictions are stored on the server and later provided to users via a dashboard.

[1612] Utilizing the Emotion Engine

[1613] server

[1614] The server uses a sentiment analyzer to analyze user sentiment in real time. For example, it analyzes customer reviews and social media posts, and if the sentiment is positive, it introduces new products or suggests cross-selling. Conversely, if the sentiment is negative, it provides customer support or sends discount coupons.

[1615] Dynamic Pricing

[1616] server

[1617] Based on the predictions, the server dynamically sets prices, for example, applying discounts to items with high inventory and raising prices for items predicted to be in high demand, aiming to maximize the company's profits.

[1618] Viewing the Dashboard

[1619] Terminal

[1620] The dashboard displays the forecast results sent from the server, using visual elements such as graphs and charts to provide information that users can easily understand. Users can also monitor the effectiveness of their marketing efforts in real time, quickly determining the success of their efforts and any necessary adjustments.

[1621] Specific examples

[1622] For example, the server collects data for the Christmas shopping season in December. First, it collects sales data from past Christmas seasons from the database, then obtains market trend data using an external API. It also collects Christmas-related data from the calendar event database. The server then imputes any missing values ​​in the collected data using the average value, scales and integrates all the data. Once all the datasets are ready, it uses them to train an LSTM model. It then performs predictions on the newly acquired data and displays the results on a dashboard. Finally, it uses an emotion engine to introduce new products if customers have positive emotions.

[1623] Prompt Sentence Examples

[1624] "For the Christmas sales season leading up to December, collect Christmas season sales data, market trend data, and seasonal event data from the past five years, and use this data to make predictions using an LSTM model and display them on a dashboard. In addition, analyze customer sentiment data in real time, and if customers show positive sentiment, introduce new products or make cross-selling suggestions."

[1625] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1626] Step 1: Data collection

[1627] server

[1628] The server first accesses the company's database and retrieves transaction data from the past five years, including details such as the quantity, price, and date of each product sold.

[1629] Input: Company database

[1630] Output: Transaction data for the past 5 years

[1631] The server then collects market trend data through external APIs, including consumer purchasing habits and competitor activity.

[1632] Input: External API

[1633] Output: Market trend data

[1634] Additionally, the server accesses a calendar events database to collect information about seasonal events such as Christmas and Valentine's Day.

[1635] Input: Calendar event database

[1636] Output: Seasonal event data

[1637] Step 2: Data Preprocessing

[1638] server

[1639] The server cleans and preprocesses all collected data. Specifically, if there are missing values ​​in the sales data, they are imputed using the mean or median. Outliers are also removed as soon as they are detected. The server also scales the data and converts it into a unified format. Finally, the sales data, market price trend data, and seasonal data are integrated into a single dataset.

[1640] Inputs: trade data, market trend data, seasonal event data

[1641] Output: Preprocessed dataset

[1642] Step 3: Train the model

[1643] server

[1644] The preprocessed dataset is used to train a machine learning model. The dataset is divided into training data and validation data. The server selects an LSTM (long short-term memory) model and inputs the data multiple times to train the model on sales demand. This training process allows the model to learn sales patterns and improve its prediction accuracy for future data.

[1645] Input: Preprocessed dataset

[1646] Output: A trained LSTM model

[1647] Step 4: Predict future sales trends

[1648] server

[1649] Using the trained LSTM model, the server analyzes newly collected data and predicts future sales trends and demand. For example, it predicts the sales volume of a particular product during the upcoming Christmas season. The prediction results are stored on the server and later provided to users via a dashboard.

[1650] Input: Newly collected data, trained LSTM model

[1651] Output: Future sales trend forecast results

[1652] Step 5: Leverage the Emotion Engine

[1653] server

[1654] The server uses a sentiment analyzer to analyze user emotions in real time. Specifically, it analyzes customer reviews and social media posts, and introduces new products and makes cross-selling suggestions to customers who show positive emotions. Conversely, it provides customer support and sends discount coupons to customers who show negative emotions.

[1655] Input: Customer reviews, social media posts

[1656] Output: Emotion-based marketing strategies

[1657] Step 6: Dynamic Pricing

[1658] server

[1659] Based on the predictions, the server dynamically sets prices, applying discounts to items with high inventory and raising prices for items predicted to be in high demand, thereby maximizing profits for the company.

[1660] Input: Future sales trend forecast results

[1661] Output: Real-time adjusted product prices

[1662] Step 7: View the dashboard

[1663] Terminal

[1664] The dashboard displays the prediction results sent from the server. The dashboard uses visual elements (graphs, charts) to provide users with easy-to-understand information. Users can also monitor the effectiveness of their marketing initiatives in real time through the dashboard, allowing them to quickly determine the success of their initiatives and any necessary adjustments.

[1665] Input: Future sales trend prediction results, emotion-based marketing strategies

[1666] Output: Visually displayed forecast results and strategies

[1667] (Application example 2)

[1668] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1669] Modern marketing strategies and sales plans must respond quickly and accurately to diverse customer needs. However, traditional systems are based solely on past sales data and market trend data, which means they cannot fully consider customer sentiment or real-time conditions. Other issues include the difficulty of improving the accuracy of demand forecasts and dynamic pricing, and providing individually optimized marketing strategies in real time. This makes it difficult for companies to optimize sales plans and inventory management, leading to a demand for systems that contribute to maximizing profits.

[1670] 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.

[1671] In this invention, the server includes means for collecting past sales data, external market trend data, and seasonal data, means for cleaning and pre-processing the collected data, means for training a machine learning model for demand forecasting using the pre-processed data, means for collecting and analyzing customer sentiment data, and means for optimizing marketing strategies based on customer sentiment, thereby improving the accuracy of demand forecasting and marketing strategies and enabling optimal sales planning and inventory management that take customer sentiment into account.

[1672] "Sales data" refers to historical data relating to past sales, inventory information, product sales quantities and prices, and the like.

[1673] "Market trend data" refers to data on customer purchasing trends, popular product trends, and economic conditions in a broad range of markets.

[1674] "Seasonal data" refers to data on fluctuations in customer purchasing behavior associated with specific seasons or events, such as Christmas or Valentine's Day.

[1675] A "machine learning model" is a collection of algorithms that learn patterns from data and make predictions or classifications based on those patterns.

[1676] "Emotional data" refers to data on the emotional movements and feelings of customers that can be obtained from their facial expressions, voice, and behavior.

[1677] A "marketing strategy" is a plan for a series of advertising, promotion, pricing, and other activities designed to achieve a specific objective.

[1678] "Training data" is a learning dataset used when building a machine learning model.

[1679] "Validation data" is a dataset used to evaluate the performance of a trained machine learning model.

[1680] A "dashboard" is an interface that visually displays forecast results and other important information in a format that is easy for users to understand.

[1681] "Dynamic pricing" is a method of adjusting product prices in real time based on demand forecasts and inventory status.

[1682] "Demand forecasting" refers to predicting future customer purchasing behavior and sales volumes based on past data and trend information.

[1683] As an embodiment of the present invention, a server uses AI and machine learning to predict future sales trends and demand, and also uses an emotion engine to generate a marketing strategy based on user emotions. Specific implementation means are described below.

[1684] Data collection

[1685] The server accesses the company's database to collect past sales data, external market trend data, and seasonal data, including the "sales data," "market trend data," and "seasonal data" previously specified. For example, the server uses an API to obtain external trend data and retrieves past sales history from the company's internal database.

[1686] Data Cleaning and Preprocessing

[1687] The collected data is cleaned, missing values ​​are filled, and outliers are removed. This is a necessary step to increase the reliability of the data. For example, missing values ​​are filled with the mean or median. The data is then scaled and sales data, market trend data, and seasonal data are integrated.

[1688] Training a machine learning model

[1689] The preprocessed data is used to train a machine learning model. The model used here is a "long short-term memory (LSTM) model" or a "neural network model." The server splits the data into training data and validation data and performs training.

[1690] Demand forecasting

[1691] The trained model is used to predict future sales trends and demand, for example, predicting demand for specific products around seasonal events like Christmas and Valentine's Day. These predictions are then displayed on a dashboard as graphs and charts that users can easily understand.

[1692] Emotion data collection and analysis

[1693] The server collects customer emotional data from the smartphone's camera and microphone. Using specialized software for emotion analysis, it analyzes facial expressions and voice in real time. This emotional data is used to optimize marketing strategies.

[1694] Optimizing your marketing strategy

[1695] Optimize your marketing strategies based on customer sentiment data. For example, if a customer expresses positive sentiment, introduce new products or offer cross-selling suggestions. If a customer expresses negative sentiment, offer customer support or send discount coupons.

[1696] Dynamic Pricing

[1697] Dynamically adjust pricing based on demand forecasts. If demand for a particular product is predicted to increase, prices are raised, while discounts are offered if inventory is high. This allows for real-time price adjustments to maximize profits.

[1698] Dashboard View

[1699] It generates a dashboard that visually displays forecast results, sentiment analysis results, and pricing information, allowing users to easily check upcoming sales plans and marketing initiatives and quickly adjust them as needed.

[1700] For example, the following prompt sentence is input to the generative AI model:

[1701] "Please forecast the demand for home video game consoles during the next Christmas season."

[1702] "Analyze recent user reviews to identify products that have received positive feedback."

[1703] This allows companies to develop highly accurate sales strategies based on market trends and customer sentiment, and implement optimized marketing strategies.

[1704] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1705] Step 1:

[1706] The server collects historical sales data from the company's database and obtains external market trend data and seasonal data through API. The input is database connection information and API key, and the output is the integrated raw data set. Specifically, it issues database queries and API requests and stores the obtained data in local storage.

[1707] Step 2:

[1708] The server performs data cleaning on the collected raw dataset. The input is the integrated raw dataset, and the output is a clean dataset. Specific operations include imputing missing values ​​(calculating the mean and median), removing outliers, and performing data scaling. This results in a dataset suitable for training machine learning models.

[1709] Step 3:

[1710] The server trains a machine learning model using a clean dataset. The input is the clean dataset, and the output is the trained model. Specifically, the data is split into training data and validation data, and training is performed using an LSTM model or a neural network model. Machine learning libraries such as TensorFlow and Keras are used to train the model.

[1711] Step 4:

[1712] The server uses the trained model to predict future sales trends and demand. The input is a newly collected dataset, and the output is the prediction result. Specifically, new data is input into the model, and predictions are made to calculate demand fluctuations and sales trends.

[1713] Step 5:

[1714] The device collects user emotional data through the smartphone's camera and microphone. The input is video and audio data, and the output is the result of emotional analysis. Specifically, it uses an emotion analysis engine to analyze changes in facial expressions and voice tone in real time to extract emotional data.

[1715] Step 6:

[1716] The server generates an individually optimized marketing strategy based on the emotional data. The input is the result of the emotional analysis, and the output is a set of marketing measures. Specifically, it runs an algorithm that performs promotions and product recommendations according to the emotional state.

[1717] Step 7:

[1718] The server performs dynamic pricing based on the demand forecast results. The inputs are the demand forecast results and inventory data, and the output is adjusted price information. Specifically, it runs an algorithm that raises prices when demand is high and applies discounts when inventory is high.

[1719] Step 8:

[1720] The terminal generates a dashboard and displays the forecast results, sentiment analysis results, and pricing information. The input is the forecast results and sentiment analysis results, and the output is a visually displayed dashboard. Specific operations include displaying information using graphs and charts, allowing users to easily check sales forecasts and marketing strategies.

[1721] Through these steps, the server and terminals work together to provide a system that realizes highly accurate demand forecasting, dynamic pricing, and emotion-based marketing strategies.

[1722] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice 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 voice data.

[1723] 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> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) 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.

[1724] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1725] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1726] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1727] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1728] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1729] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1730] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1731] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1732] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1733] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1734] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1735] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1736] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1737] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1738] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1739] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1740] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1741] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1742] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1743] The following is further disclosed regarding the above embodiment.

[1744] (Claim 1)

[1745] a means for collecting historical sales data, external market trend data and seasonality data;

[1746] means for cleaning and pre-processing the collected data;

[1747] means for training a machine learning model for demand forecasting using the preprocessed data; and

[1748] a means for forecasting future sales trends and demand using the trained model;

[1749] means for generating a dashboard for displaying the prediction results;

[1750] A means for dynamically setting prices based on the results of the predictions;

[1751] A means of generating individually optimized marketing strategies based on customer data;

[1752] A system including:

[1753] (Claim 2)

[1754] The system according to claim 1, which performs missing value completion and outlier removal based on the collected data.

[1755] (Claim 3)

[1756] 2. The system of claim 1, wherein the preprocessed data is divided into training data and validation data, and the training data is used to train the time series forecasting model or the neural network model.

[1757] "Example 1"

[1758] (Claim 1)

[1759] a means for collecting historical sales data, external market trend data and seasonality data;

[1760] means for cleaning and pre-processing the collected data;

[1761] means for training a machine learning model for demand forecasting using the preprocessed data; and

[1762] a means for forecasting future sales trends and demand using the trained model;

[1763] means for generating a recording medium for displaying the prediction result;

[1764] A means for dynamically setting prices based on the results of the predictions;

[1765] A means of generating individually optimized marketing strategies based on customer data;

[1766] A system including:

[1767] (Claim 2)

[1768] The system according to claim 1, which performs missing value completion and outlier removal based on the collected data.

[1769] (Claim 3)

[1770] 2. The system of claim 1, wherein the preprocessed data is divided into training data and validation data, and the training data is used to train the time series forecasting model or the neural network model.

[1771] "Application Example 1"

[1772] (Claim 1)

[1773] a means for collecting historical sales data, external market trend data and seasonality data;

[1774] means for cleaning and pre-processing the collected data;

[1775] means for training a machine learning model for demand forecasting using the preprocessed data; and

[1776] a means for forecasting future sales trends and demand using the trained model;

[1777] means for generating a dashboard for displaying the prediction results;

[1778] A means for dynamically setting prices based on the results of the predictions;

[1779] A means of generating individually optimized marketing strategies based on customer data;

[1780] means for providing a smartphone application that performs real-time demand forecasting and dynamic pricing;

[1781] A system including:

[1782] (Claim 2)

[1783] The system according to claim 1, which performs missing value completion and outlier removal based on the collected data.

[1784] (Claim 3)

[1785] 2. The system of claim 1, wherein the preprocessed data is divided into training data and validation data, and the training data is used to train the time series forecasting model or the neural network model.

[1786] (Claim 4)

[1787] 10. The system of claim 1, comprising a smartphone application for real-time demand forecasting and dynamic pricing.

[1788] "Example 2: Combining Emotion Engines"

[1789] (Claim 1)

[1790] a means for collecting historical transaction data, external market trend data and seasonality information;

[1791] means for cleaning and pre-processing the collected data;

[1792] means for training a machine learning model for demand forecasting using the preprocessed data; and

[1793] a means for forecasting future sales trends and demand using the trained model;

[1794] means for generating a visualization interface for displaying the prediction results;

[1795] A means for dynamically setting prices based on the results of the predictions;

[1796] means for analyzing customer emotion data using a sentiment analysis device and generating individually optimized marketing strategies based thereon;

[1797] A system including:

[1798] (Claim 2)

[1799] The system according to claim 1, which performs missing value completion and outlier removal based on the collected data.

[1800] (Claim 3)

[1801] 10. The system of claim 1, wherein the preprocessed data is divided into training data and validation data, and the training data is used to train the time series forecasting model or the artificial neural network model.

[1802] "Application example 2 when combining emotion engines"

[1803] (Claim 1)

[1804] a means for collecting historical sales data, external market trend data and seasonality data;

[1805] means for cleaning and pre-processing the collected data;

[1806] means for training a machine learning model for demand forecasting using the preprocessed data; and

[1807] a means for forecasting future sales trends and demand using the trained model;

[1808] means for generating a dashboard for displaying the prediction results;

[1809] A means for dynamically setting prices based on the results of the predictions;

[1810] A means of generating individually optimized marketing strategies based on customer data;

[1811] means for collecting and analyzing customer sentiment data;

[1812] A means to optimize marketing strategies based on customer sentiment;

[1813] A system including:

[1814] (Claim 2)

[1815] The system according to claim 1, which performs missing value completion and outlier removal based on the collected data.

[1816] (Claim 3)

[1817] 2. The system of claim 1, wherein the preprocessed data is divided into training data and validation data, and the training data is used to train the time series forecasting model or the neural network model.

[1818] (Claim 4)

[1819] 2. The system according to claim 1, further comprising means for collecting customer emotion data using a camera and a microphone and analyzing the data with an emotion analysis model.

[1820] (Claim 5)

[1821] 10. The system of claim 1, further comprising means for adjusting prices in real time based on dynamic pricing to maximize sales. [Explanation of symbols]

[1822] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. a means for collecting historical sales data, external market trend data and seasonality data; means for cleaning and pre-processing the collected data; means for training a machine learning model for demand forecasting using the preprocessed data; and a means for forecasting future sales trends and demand using the trained model; means for generating a dashboard for displaying the prediction results; A means for dynamically setting prices based on the results of the predictions; A means of generating individually optimized marketing strategies based on customer data; A system including:

2. The system according to claim 1, wherein missing values ​​are imputed and outliers are removed based on the collected data.

3. 2. The system of claim 1, wherein the preprocessed data is divided into training data and validation data, and the training data is used to train the time series prediction model or the neural network model.

Images