System

Abstract

A system is provided.SOLUTION: A system comprising: means for collecting performance data of past advertisements; means for preprocessing the collected data; means for performing AI analysis based on the preprocessed data; means for extracting an improvement point of an advertisement creative based on a result of the AI analysis; means for generating a report including the improvement point; and means for providing the generated report to a terminal of a user.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] Traditional advertising creative production relies primarily on the subjective judgment and experience of designers and marketers, making it difficult to efficiently improve. This often results in inconsistent ad performance and a lack of specific metrics for producing statistically superior ads. This can lead to lower ad click-through rates and conversion rates, leading to concerns about a worsening return on advertising investment (ROI). [Means for solving the problem]

[0005] This invention provides a means for collecting performance data of past advertisements and preprocessing the data. It also includes a means for performing AI analysis based on the preprocessed data. It also provides a means for extracting specific improvement points for advertising creatives based on the results of this analysis. It also includes a means for generating a report including the extracted improvement points and providing the report to the user's device, thereby achieving efficient improvement of advertising creatives and increased advertising efficiency.

[0006] "Advertising creative" refers to the visual and textual elements created for advertising purposes, including images, videos, headlines, taglines, etc.

[0007] "Performance data" refers to data that measures how a particular advertisement has performed in the past, and specifically includes metrics such as click-through rate, conversion rate, and number of impressions.

[0008] "Means of collection" refers to the programs and protocols used to obtain performance data from advertising delivery platforms and databases.

[0009] "Preprocessing means" refers to processing methods used to prepare collected data for analysis by filling in missing values, filtering, and normalizing the data scale.

[0010] "Means for performing AI analysis" refers to a program that uses machine learning algorithms to analyze pre-processed data and predict or evaluate the effectiveness of advertising creatives.

[0011] "Means for extracting points for improvement" refers to a means of specifically indicating which elements of advertising creative should be improved and how, based on the results of AI analysis.

[0012] "Means for generating reports" refers to a program that visualizes the extracted improvement points and creates a report in an easy-to-understand format using charts and graphs.

[0013] The "means for providing" refers to a communication means or display means for transmitting the generated report to the user's terminal so that the user can check it. [Brief explanation of the drawings]

[0014] [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 showing 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

[0015] 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.

[0016] First, the terms used in the following description will be explained.

[0017] 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).

[0018] 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.

[0019] 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.

[0020] 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.

[0021] 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."

[0022] [First embodiment]

[0023] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0024] 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.

[0025] 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).

[0026] 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.

[0027] 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.

[0028] 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.

[0029] 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.

[0030] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0031] 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.

[0032] 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.

[0033] 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.

[0034] 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."

[0035] The present invention is a system for efficiently improving advertising creatives, and is specifically implemented as follows: The system is mainly composed of components of a server, a terminal, and a user.

[0036] Data collection and preprocessing

[0037] In the initial stage, the server collects historical ad performance data from ad serving platforms and databases. This data includes performance indicators such as click-through rates, conversion rates, and number of impressions. Once collected, the server fills in missing values, filters, and normalizes the data to prepare it for analysis. This ensures consistency and accuracy of the data.

[0038] AI analysis

[0039] The server then uses machine learning algorithms to analyze the pre-processed data. Learning models include random forests and neural networks, and evaluate the impact that specific elements of ad creative have on performance indicators. This analysis allows the system to learn the characteristics of past successful and unsuccessful ads, making it possible to predict the performance of new ads.

[0040] Extracting points for improvement

[0041] Based on the results of AI analysis, the server extracts specific points for improvement in the ad creative. For example, if it concludes that changes in color, text font, or layout will contribute to an increase in click-through rate, it presents that information as a specific improvement suggestion, providing designers and marketers with clear guidelines.

[0042] Generate and deliver reports

[0043] The server visually organizes the identified improvement points and generates a report using charts and graphs. The report is output in PDF or web-based dashboard format and is designed for easy viewing by the user. Finally, the server sends the generated report to the terminal, where the user can review it and take improvement actions.

[0044] Specific examples

[0045] For example, let's say an e-commerce site has a very successful summer sales campaign with the following performance data:

[0046] Click-through rate: 5%

[0047] Conversion rate: 2%

[0048] Views: 100,000

[0049] In contrast, here's the data for a failed fall sales campaign ad:

[0050] Click-through rate: 1%

[0051] Conversion rate: 0.5%

[0052] Views: 80,000

[0053] The server collects this data and performs preprocessing. The AI ​​analysis reveals that bright colors and highly visible fonts were effective for summer advertising. The extracted improvement points are compiled into a report as specific suggestions, such as "brighten the colors of advertising images and change the text font to a larger, more visible one."

[0054] Finally, the server provides this report to the user's device, and the user can design the next advertising campaign based on the suggested improvements, which can improve advertising efficiency. Through this process, an improvement in advertising return on investment (ROI) can be realized.

[0055] The processing flow will be explained below.

[0056] Step 1:

[0057] The server sends an API request to an ad serving platform or database to collect performance data for past ads, including performance metrics such as click-through rates, conversion rates, and impressions.

[0058] Step 2:

[0059] The server preprocesses the collected data by imputing missing values, filtering for consistency, and normalizing the data scale, making the data suitable for analysis.

[0060] Step 3:

[0061] The server then feeds the pre-processed data into machine learning algorithms, such as models like random forests or neural networks, that analyze the impact of each element of the ad creative on performance.

[0062] Step 4:

[0063] The server trains the machine learning model and checks the model's predictive accuracy through cross-validation, thereby building an accurate predictive model.

[0064] Step 5:

[0065] The server extracts specific points for improvement based on the analysis results. For example, it identifies common characteristics of ads with high click-through rates and conversion rates and summarizes them as areas for improvement. This information is presented in the form of suggestions for color, font, layout, etc.

[0066] Step 6:

[0067] The server visually organizes the identified improvement points and generates a report using charts and graphs. The report is created in PDF or web-based dashboard format, with an emphasis on ease of viewing and understanding.

[0068] Step 7:

[0069] The server sends the generated report to the user's terminal, which allows the user to quickly take concrete action regarding the creation of new advertisements.

[0070] Step 8:

[0071] We review the reports received by the user and design and produce new ad creatives based on the suggested improvements. Based on the suggestions, we make specific improvements such as changes to color usage, layout, and text fonts.

[0072] Step 9:

[0073] The user then redistributes the improved ad, collecting new performance data. This data is then sent back to the server for analysis in the next cycle. By repeating this process, ad performance continues to improve.

[0074] Example 1

[0075] 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."

[0076] Improving the performance of advertising creatives is important in marketing, but conventional methods have made it difficult to quickly and accurately identify effective areas for improvement. To solve this problem, there is a need to develop a system that can efficiently and accurately extract and suggest areas for improvement using past advertising performance data.

[0077] 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.

[0078] In this invention, the server includes a means for collecting performance data of past advertisements, a means for preprocessing the collected data, and a means for performing AI analysis based on the preprocessed data, which makes it possible to evaluate the impact of elements of advertising creatives on performance indicators and quickly and accurately extract areas for improvement.

[0079] "Past advertising performance data" refers to data that includes advertising performance indicators such as click rates, conversion rates, and number of impressions.

[0080] "Means of collection" refers to the means for obtaining past advertising performance data from advertising distribution platforms and databases.

[0081] The "preprocessing means" refers to means for imputing missing values ​​in the collected data, filtering the data, and normalizing the data.

[0082] "Means for conducting AI analysis" means means for applying machine learning algorithms using pre-processed data to evaluate the impact of elements of advertising creative on performance metrics.

[0083] "Means for extracting points for improvement in advertising creative" refers to a means for identifying specific changes to improve the effectiveness of advertising based on the results of AI analysis.

[0084] The "means for generating a report" refers to a means for visually organizing the extracted improvement points and creating a report using charts and graphs.

[0085] The "means for providing to the user's terminal" refers to the means for providing the generated report in a form that can be viewed by the user, such as a PDF format or a web-based dashboard.

[0086] A "data analysis algorithm" is a mathematical model or computational method for finding patterns and trends in data and making predictions.

[0087] "Data reconciliation" is the process of converting data collected from different data sources into a unified format and making it consistent.

[0088] "Data standardization" is the process of scaling all numerical data to a certain range, typically normalizing it to the range 0 to 1.

[0089] This invention is a system aimed at efficiently improving advertising creatives. The system is mainly composed of components such as a server, a terminal, and a user.

[0090] Data collection and preprocessing

[0091] The server collects past advertising performance data from advertising distribution platforms and databases. This data includes performance indicators such as click-through rates, conversion rates, and number of impressions. A specific example of how this data can be collected is by using the Google (registered trademark) Ads API. The API is used to obtain advertising performance data for a specified period and store it in a database.

[0092] The server performs missing value imputation, filtering, and normalization on the collected data. This preprocessing is performed using the Python Pandas library. Specifically, the following operations are performed:

[0093] Use the mean or median to impute missing values.

[0094] To remove unnecessary data, filtering is performed according to specific conditions.

[0095] Scale all numeric data to the range 0 to 1.

[0096] AI analysis

[0097] The server runs machine learning algorithms on the preprocessed data, such as random forests and neural networks. Specifically, it uses the Scikit-learn library to train a random forest model to predict ad performance.

[0098] The server performs the following operations:

[0099] Using the Scikit-learn library, training data is input into the model and learning is performed.

[0100] Use the trained model to predict performance on new advertising data.

[0101] Extracting points for improvement

[0102] Based on the results of the AI ​​analysis, the server extracts specific points for improvement in the ad creative. From the analysis results, it identifies the factors that have the greatest impact on performance and forms specific improvement proposals based on those.

[0103] Specific suggestions include using bright colors and highly visible fonts. This information is provided as useful feedback to designers and marketers.

[0104] Generate and deliver reports

[0105] The server visually organizes the identified improvement points and generates a report using charts and graphs. This report is output in PDF format and a web-based dashboard format. The Matplotlib library is used to generate various graphs and convert them to PDF.

[0106] The server finally delivers this report to the user's device, either by emailing the generated PDF file or uploading it to a web dashboard where the user can log in and view it.

[0107] Specific examples

[0108] For example, let's say an e-commerce site runs a successful summer sales campaign with the following performance data:

[0109] Click-through rate: 5%

[0110] Conversion rate: 2%

[0111] Views: 100,000

[0112] In contrast, here's the data for a failed fall sales campaign ad:

[0113] Click-through rate: 1%

[0114] Conversion rate: 0.5%

[0115] Views: 80,000

[0116] The server collects this data and performs preprocessing. The AI ​​analysis reveals that bright colors and highly visible fonts were effective for summer advertising. The extracted improvement points are compiled into a report as specific suggestions, such as "brighten the colors of advertising images and change the text font to a larger, more visible one."

[0117] Prompt Sentence Examples

[0118] Users are prompted to take specific actions, such as, "Use bright colors and large, highly visible text fonts in your next ad campaign images to increase click-through rates and conversion rates."

[0119] This series of systems makes it possible to improve advertising return on investment (ROI).

[0120] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0121] Step 1: Data collection

[0122] The first thing the server does is collect past advertising performance data from ad distribution platforms and databases. Specifically, it uses the Google Ads API to obtain data such as click-through rates, conversion rates, and number of impressions for a specified period. The input data is the unprocessed raw data obtained from the API, and the output is past advertising performance data stored in the server's database.

[0123] Specific behavior:

[0124] Access the Google Ads API using an authentication token to retrieve the required data.

[0125] The acquired data is saved in a CSV file or database.

[0126] Step 2: Data Preprocessing

[0127] The server then preprocesses the collected data, which includes missing value imputation, data filtering, and normalization. The input to the preprocessing is the raw data collected, and the output is a consistent and accurate dataset that can be analyzed.

[0128] Specific behavior:

[0129] Create a data frame using Python's Pandas library.

[0130] Impute missing values ​​with the mean or median.

[0131] Filter data based on specific criteria to remove unwanted data.

[0132] Scales numeric data to the range 0 to 1.

[0133] Step 3: AI analysis

[0134] The server runs a machine learning algorithm on the preprocessed data. This step uses models like random forests or neural networks. The input is the preprocessed dataset, and the output is an analysis showing the impact of each element of the ad creative on performance metrics.

[0135] Specific behavior:

[0136] Train a random forest model using the Scikit-learn library.

[0137] Use the trained model to predict performance on new advertising data.

[0138] Step 4: Identifying areas for improvement

[0139] The server extracts specific points for improvement in the ad creative based on the results of the AI ​​analysis. The input is the AI ​​analysis results, and the output is specific improvement suggestions.

[0140] Specific behavior:

[0141] The factors with the greatest impact are extracted from the analysis results.

[0142] Create specific improvement suggestions for the use of colors in advertising images and text fonts.

[0143] Step 5: Generate reports

[0144] The server visually organizes the extracted improvement points and generates a report using charts and graphs. The input is the improvement proposals, and the output is a report in PDF format or a web dashboard format.

[0145] Specific behavior:

[0146] Generate various graphs using the Matplotlib library.

[0147] Convert the generated graphs to PDF and compile a report.

[0148] Step 6: Providing the report

[0149] The server provides the generated report to the user's device. This step involves emailing the PDF file or uploading it to a web dashboard. The input is the generated report and the output is the report provided to the user.

[0150] Specific behavior:

[0151] Generate a PDF report and send it to the user via email.

[0152] Upload reports to a web dashboard so users can log in and view them.

[0153] (Application example 1)

[0154] 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."

[0155] Conventional methods for improving advertising creatives require time and effort to collect and analyze past performance data, making it difficult to propose effective improvements in real time. Furthermore, since specific improvement proposals based on the expected effectiveness of advertising creatives cannot be obtained, there are limited ways to maximize advertising performance.

[0156] 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.

[0157] In this invention, the server includes means for collecting performance data of past advertisements, means for preprocessing the collected data, means for performing AI analysis based on the preprocessed data, means for extracting improvement points for the advertisement creative based on the results of the AI ​​analysis, means for generating a report including the improvement points, means for providing the generated report to a user's terminal, means for providing improvement suggestions for the advertisement creative in real time, and means for generating responsive prompt sentences. This enables efficient creative improvement in real time based on the performance data of advertisements, thereby maximizing advertisement performance.

[0158] "Past advertising performance data" refers to data including performance indicators related to past advertising distribution, such as click rates, conversion rates, and number of impressions of advertisements.

[0159] "Data preprocessing" is the process of completing the collected data by imputing missing values, filtering, and normalizing it, preparing it for analysis.

[0160] "AI analysis" is the process of analyzing data using machine learning algorithms to assess the impact that specific elements of ad creative have on performance metrics.

[0161] "Improvement points" are specific suggested changes to improve the performance of ad creatives, extracted based on the results of AI analysis.

[0162] "Report generation" is the process of visually organizing improvement points and outputting them in a format that is easy for users to understand using charts and graphs.

[0163] "User device" refers to an electronic device owned by a user, such as a smartphone, tablet, or PC, to which a report is provided.

[0164] "Means for providing real-time improvement suggestions for advertising creatives" is a function for quickly providing users with improvement suggestions for advertising creatives that are generated based on the results of AI analysis.

[0165] The "means for generating prompt sentences" is a function that automatically creates sentences to make suggestions or ask questions to the user in order to support the improvement of advertising creatives.

[0166] This invention is a system for realizing efficient improvement of advertising creatives, and is composed of the following components: The system includes the following elements: a server, a terminal, and a user.

[0167] Data collection and preprocessing

[0168] In the initial stage, the server collects historical ad performance data from ad serving platforms and databases. This data includes performance indicators such as click-through rates, conversion rates, and number of impressions. Once collected, the server fills in missing values, filters, and normalizes the data to prepare it for analysis. This ensures consistency and accuracy of the data.

[0169] AI analysis

[0170] After data preprocessing is complete, the server performs analysis using machine learning algorithms. Learning models used include random forests and neural networks. This allows the impact of specific elements of ad creative on performance indicators to be evaluated. AI analysis learns the characteristics of past successful and unsuccessful ads, allowing for prediction of the performance of new ads.

[0171] Extracting points for improvement

[0172] Based on the results of the AI ​​analysis, the server extracts specific points for improvement in the ad creative. For example, if it concludes that changes in color, text font, or layout will contribute to an increase in click-through rate, it will present that information as a specific improvement suggestion.

[0173] Generate and deliver reports

[0174] The server visually organizes the extracted improvement points and generates a report using charts and graphs. The report is output in PDF or web-based dashboard format and is designed to be easy for users to view. Finally, the server sends the generated report to the terminal, where the user can review it and take improvement actions.

[0175] Real-time improvement suggestions for advertising creatives

[0176] Based on AI analysis, the server generates real-time suggestions for improving ad creatives and provides them to users' devices. When users create new ads, they can make specific changes while receiving immediate feedback from the server.

[0177] Generate prompt statement

[0178] Furthermore, the server can generate prompt sentences to assist the user in creating an advertisement. The prompt sentences are sentences that suggest improvements to the advertisement creative and serve to prompt the user to take specific action to improve the advertisement.

[0179] Specific examples

[0180] For example, an e-commerce site may have a very successful summer sales campaign with the following performance data:

[0181] Click-through rate: 5%

[0182] Conversion rate: 2%

[0183] Views: 100,000

[0184] In contrast, here's the data for a failed fall sales campaign ad:

[0185] Click-through rate: 1%

[0186] Conversion rate: 0.5%

[0187] Views: 80,000

[0188] The server collects this data and performs preprocessing. The AI ​​analysis reveals that bright colors and highly visible fonts were effective for summer advertising. The extracted improvement points are compiled into a report as specific suggestions, such as "brighten the colors of advertising images and change the text font to a larger, more visible one." Finally, the server provides this report to the user's device, and the user can design the next advertising campaign based on the suggested improvements, which is expected to improve advertising efficiency.

[0189] Also, when a user creates a new ad creative, the server provides a prompt like this:

[0190] "We recommend using brighter colors and more legible fonts in your new ads. This change may increase your click-through rate."

[0191] This allows the system to help improve effective advertising creative in real time.

[0192] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0193] Step 1:

[0194] The server collects performance data of past ads from ad distribution platforms and databases. The collected data includes click-through rates, conversion rates, number of impressions, etc. The input is the performance data of the ads, and the output is the collected raw data.

[0195] Step 2:

[0196] The server preprocesses the collected data. Specifically, it imputes missing values ​​in the data, filters to remove noise, and scales and normalizes the data. The input is the raw data collected in step 1, and the output is the preprocessed data.

[0197] Step 3:

[0198] The server uses the preprocessed data to perform AI analysis, using machine learning models such as random forests and neural networks to evaluate the impact of specific elements of the ad creative on performance metrics. The input is the preprocessed data, and the output is the analysis results.

[0199] Step 4:

[0200] The server extracts specific points for improvement in the ad creative based on the results of the AI ​​analysis. For example, it generates specific suggestions, such as how changing the color usage or text font will contribute to improving click-through rates. The input is the analysis results obtained in step 3, and the output is the points for improvement.

[0201] Step 5:

[0202] The server generates a visually organized report using the extracted improvement points. This report is output in a user-friendly format using charts and graphs. The input is the improvement points extracted in step 4, and the output is the report.

[0203] Step 6:

[0204] The server provides the generated report to the user's device, allowing the user to review the report and take specific improvement actions. The input is the report generated in step 5, and the output is the report delivered to the user's device.

[0205] Step 7:

[0206] The server provides real-time suggestions for improving ad creatives. Users can receive immediate feedback when creating new ad creatives. The input is the initial data for new ad creatives and analysis results based on past performance data, and the output is real-time suggestions for improvement.

[0207] Step 8:

[0208] The server generates a prompt to assist the user in creating an ad. The prompt serves to encourage the user to take specific action to improve the ad. The input is the improvement proposal obtained in step 7, and the output is the prompt. A specific example would be, "We recommend that you use brighter colors in your new ad and change the font to one that is more visible. This change may improve your click-through rate."

[0209] 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.

[0210] This invention is a system aimed at efficiently improving advertising creatives, utilizing both past advertising performance data and user emotional data. The system is primarily comprised of a server, a terminal, and a user. Here, we will explain in detail how this system is implemented.

[0211] Data collection and preprocessing

[0212] At the initial stage, the server sends an API request to an ad serving platform or database to collect performance data of past ads. The collected data includes performance indicators such as click-through rate, conversion rate, and number of impressions. In addition, the server uses an emotion engine to obtain emotional data from the user's facial expressions, voice tone, text input, etc. This also collects the emotions of the user viewing the ad.

[0213] Data Preprocessing

[0214] The server preprocesses the advertising performance data and sentiment data collected. Specifically, it fills in missing values ​​in the data and performs filtering to maintain consistency. It also normalizes the data scale and prepares it for analysis. This makes both data suitable for analysis.

[0215] AI analysis

[0216] The server uses machine learning algorithms to analyze the preprocessed advertising performance data and emotion data. The learning models include random forests and neural networks, and evaluate the impact of each element of the advertising creative on performance indicators and user emotions. This analysis learns the characteristics of past successful and unsuccessful ads, making it possible to predict the performance and evaluate emotions of new ads.

[0217] Extracting points for improvement

[0218] Based on the results of the AI ​​analysis, the server extracts specific areas for improvement in the ad creative. For example, it identifies common characteristics of ads with high click-through rates and conversion rates, as well as the corresponding positive emotional responses, and summarizes these as areas for improvement. This information is presented in the form of suggestions for color usage, fonts, layout, etc.

[0219] Generate and deliver reports

[0220] The server visually organizes the extracted improvement points and generates a report using charts and graphs. The report is created in PDF or web-based dashboard format, emphasizing ease of viewing and understanding. It also includes the results of sentiment data analysis, showing how advertising affects users' emotions. Finally, the server provides the generated report to the user's device.

[0221] Specific examples

[0222] For example, let's say an e-commerce site's summer sales campaign was extremely successful, resulting in the following performance and sentiment data:

[0223] Click-through rate: 5%

[0224] Conversion rate: 2%

[0225] Views: 100,000

[0226] Positive emotional responses (smiles, positive comments): 70%

[0227] In contrast, here's the data for a failed fall sales campaign ad:

[0228] Click-through rate: 1%

[0229] Conversion rate: 0.5%

[0230] Views: 80,000

[0231] Negative emotional responses (sad facial expressions, negative comments): 50%

[0232] The server collects this data and performs preprocessing. The AI ​​analysis results reveal that bright colors and highly visible fonts are effective in summer advertisements and also evoke positive emotional responses from users. The extracted improvement points are compiled into a report as specific suggestions, such as "brighten the colors of the advertisement images and change the text font to a larger, more visible one."

[0233] Finally, the server provides this report to the user's device, and the next advertising campaign is designed based on the user's suggested improvements, which is expected to improve advertising efficiency. Through this series of processes, advertising return on investment (ROI) and advertising that takes user emotions into consideration are realized.

[0234] The processing flow will be explained below.

[0235] Step 1:

[0236] The server sends an API request to an ad serving platform or database to collect performance data of past ads. The collected data includes performance indicators such as click-through rate, conversion rate, and number of impressions. At the same time, an emotion engine is run to obtain emotional data from the user's facial expressions, tone of voice, text input, etc.

[0237] Step 2:

[0238] The server preprocesses the advertising performance data and sentiment data collected. First, if any missing values ​​exist, they are imputed appropriately. Next, the data is filtered to ensure consistency. Finally, the scale of the input data is normalized and converted into a format suitable for machine learning algorithms.

[0239] Step 3:

[0240] The server then feeds the pre-processed data into machine learning algorithms, such as random forests and neural networks, to evaluate the impact of each element of the ad creative on performance metrics and user sentiment.

[0241] Step 4:

[0242] The server trains the machine learning model and performs cross-validation, checking the accuracy of the model based on the results of the cross-validation and optimizing the model as needed, thereby building a more accurate predictive model.

[0243] Step 5:

[0244] Based on the analysis results, the server extracts specific points for improvement in the ad creative. For example, it identifies common characteristics of ads with high click-through rates and conversion rates, as well as elements that evoke positive emotional responses. This information is presented as improvement suggestions in the form of specific things like color usage, fonts, and layout.

[0245] Step 6:

[0246] The server visually organizes the identified improvement points and generates a report using charts and graphs in PDF or web-based dashboard format, which also includes the results of sentiment data analysis to make it easier to understand the overall picture.

[0247] Step 7:

[0248] The server sends the generated report to the user's device, allowing the user to review the report and use it to design and improve new ad creatives.

[0249] Step 8:

[0250] We carefully review the reports received by users and design and produce new ad creatives based on the suggested improvements. Based on the suggestions, we make specific improvements such as changes to color usage, layout, and text fonts.

[0251] Step 9:

[0252] The user then redistributes the improved ad, collecting new performance data. This new data is then sent back to the server for analysis in the next cycle. This process is repeated, resulting in continuous improvements in ad performance and user emotional response.

[0253] Example 2

[0254] 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."

[0255] Conventional advertising improvement systems have been unable to effectively improve advertising performance because they have been unable to fully integrate past advertising performance data and user emotion data, and their specific proposals for improving advertising creatives are abstract. Furthermore, it has been difficult to efficiently collect and analyze user emotion data and perform emotion evaluation based on advertising performance.

[0256] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting past advertising performance data, means for pre-processing the collected data, means for collecting user emotion data, means for performing AI analysis based on the pre-processed data, means for extracting improvement points for advertising creative based on the results of the AI ​​analysis, means for generating a report including the improvement points, and means for providing the generated report to the user's terminal. This makes it possible to comprehensively analyze past advertising performance data and user emotion data and effectively improve advertising performance through specific improvement suggestions.

[0257] "Advertising performance data" refers to performance indicators such as click rates, conversion rates, and number of impressions that indicate the effectiveness of past advertising campaigns.

[0258] "Preprocessing" refers to the process of filling in missing values ​​in the data, filtering to maintain consistency, and normalizing the data to prepare it for analysis.

[0259] "User emotion data" refers to data that indicates the user's emotional state, obtained from the user's facial expression, voice tone, text input, and the like.

[0260] "AI analytics" is the process of using machine learning algorithms to analyze pre-processed data and evaluate the impact of each element of advertising creative on performance metrics and user sentiment.

[0261] "Points for improving advertising creative" refers to specific changes and improvements that are needed to improve advertising performance based on the results of AI analysis.

[0262] A "report" is a report that visually organizes the analysis results and extracted improvement points and displays them using charts, graphs, etc.

[0263] A "user's terminal" is a device such as a computer, smartphone, or tablet that a user uses to receive reports provided by the server.

[0264] This invention is a system aimed at efficiently improving advertising creatives, and provides a means for integrating and analyzing past advertising performance data and user emotion data. The system is primarily comprised of components: a server, a terminal, and a user. Here, we will explain in detail how this system is implemented.

[0265] Data collection

[0266] The server first sends an API request to an ad serving platform or database to collect past ad performance data. Specific APIs typically used are the Google Ads API and Facebook Ads API. This data includes click-through rates, conversion rates, and number of impressions.

[0267] The server also collects the user's emotional data using an emotion engine. The emotion engine uses, for example, Microsoft® Azure® Emotion API or IBM Watson® Tone Analyzer, which analyzes and obtains the user's emotional state from their facial expressions, voice tone, and text input. Emotion data is collected when the device captures the user's video and audio and sends them to the server.

[0268] Data Preprocessing

[0269] The server preprocesses the advertising performance data and sentiment data collected by the server. The Python pandas library is used to impute missing values, and data filtering and normalization are also performed. For example, this includes outlier detection and normalization of the data scale using a scaling library.

[0270] AI analysis

[0271] The server then analyzes the pre-processed data using machine learning algorithms, specifically models such as random forests and neural networks. The analysis uses the scikit-learn library to train and evaluate the models. The AI ​​analysis evaluates the impact of each element of the ad creative on performance metrics and user sentiment.

[0272] Extracting points for improvement

[0273] The server extracts specific areas for improvement in the ad creative based on the results of the AI ​​analysis. It identifies common characteristics of ads with high click-through rates and conversion rates and summarizes the areas for improvement. For example, it generates suggestions for "bright colors" and "highly visible fonts."

[0274] Generate and deliver reports

[0275] The server visually organizes the extracted improvement points and generates a report using charts and graphs. Matplotlib and Plotly are used to generate the report, which is automatically created using a PDF generator. Finally, the server provides the generated report to the user's device. The report is delivered as a download link or email attachment, and the user can use it to design their next advertising campaign.

[0276] Specific examples

[0277] For example, suppose an e-commerce site runs a summer sale campaign ad with a 5% click-through rate, a 2% conversion rate, 100,000 impressions, and a 70% positive emotional response. In contrast, if a fall sale campaign ad has a 1% click-through rate, a 0.5% conversion rate, 80,000 impressions, and a 50% negative emotional response, the server collects and preprocesses this data and performs AI analysis. The analysis results suggest that bright colors and highly visible fonts are effective for summer ads.

[0278] The extracted improvement points are compiled into a report with specific suggestions, such as "Use brighter colors in the ad images and change the text font to a larger, more visible one." Finally, the server provides the generated report to the user's device, allowing the user to incorporate the improvements into their next ad campaign.

[0279] Prompt Sentence Examples

[0280] Please suggest areas for improvement in the ad creative based on the advertising data below.

[0281] (Successful summer sale campaign advertising data)

[0282] Click-through rate: 5%

[0283] Conversion rate: 2%

[0284] Views: 100,000

[0285] Positive emotional response: 70%

[0286] (Failed Fall Sale Campaign Advertising Data)

[0287] Click-through rate: 1%

[0288] Conversion rate: 0.5%

[0289] Views: 80,000

[0290] Negative emotional reactions: 50%

[0291] suggestion:

[0292] Use bright colors and highly visible fonts, just like in summer ads, to increase click-through rates and conversion rates.

[0293] With the above configuration, this system improves advertising performance and proposes advertising creatives that take user emotions into consideration.

[0294] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0295] Step 1: Collect advertising performance data

[0296] The server sends API requests to ad distribution platforms and databases to collect past ad performance data. Specifically, it uses the Google Ads API and Facebook Ads API. The input is API authentication information and the query required to obtain data, and the output is data on ad performance indicators such as click-through rate (CTR), conversion rate (CVR), and number of impressions. The server performs API authentication using the client ID and secret, and obtains data from the specified endpoint.

[0297] Step 2: Collecting emotion data

[0298] The server runs an emotion engine and collects emotion data from the user's facial expressions, voice tone, and text input. The inputs are camera footage, audio clips, and text data, and the emotional state is analyzed based on this. Specifically, the Microsoft Azure Emotion API and IBM Watson Tone Analyzer are used. The device captures the user's video and audio and sends them to the server to obtain emotion data. The output is data indicating the user's emotional state (e.g., smile rate, negative comment rate).

[0299] Step 3: Preprocessing the data

[0300] The advertising performance data and sentiment data collected by the server are preprocessed. The input is the raw data obtained in steps 1 and 2. Specific preprocessing includes missing value imputation, filtering, and normalization. The Python pandas library is used to impute missing values ​​with the mean or median, and outliers are filtered. Furthermore, a scaling library is used to normalize the data. The output is data that is ready for analysis.

[0301] Step 4: AI analysis

[0302] The server applies a machine learning algorithm to the preprocessed data and performs analysis. The input is the preprocessed advertising performance data and emotion data. Specifically, a random forest or neural network model is used, and the model is trained and evaluated using the scikit-learn library. The output is an evaluation of the impact of each element of the advertising creative on performance indicators and user emotions. The server randomly shuffles the training data and divides it into a training set and a test set. The algorithm is applied to train the model, and accuracy is verified on the test set.

[0303] Step 5: Identifying areas for improvement

[0304] The server extracts specific improvement points for the ad creative based on the results of the AI ​​analysis. The input is the data results of the AI ​​analysis. The server analyzes the importance of the features and displays a ranking of the elements that have the greatest impact on performance. The improvement points are written out as specific suggestions using natural language generation (NLG) technology. The output is specific changes and improvements (e.g., bright colors, highly visible fonts) as improvement suggestions.

[0305] Step 6: Generate reports

[0306] The server visually organizes the improvement points extracted and generates a report using charts and graphs. The input is improvement point data. Matplotlib and Plotly are used to generate the report. A data visualization library is used to graphically represent each indicator, and a PDF generator is used to automatically generate the report. The output is a report in PDF format or a web-based dashboard format.

[0307] Step 7: Providing the report

[0308] The server provides the generated report to the user's device. The input is the generated report data. The server configures the SMTP settings to send the report to the user's email address, and includes links and attachments in the email body. The output is a report provided to the user's device, which the user can use to design their next advertising campaign.

[0309] (Application example 2)

[0310] 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."

[0311] Conventional systems for improving advertising creatives rely solely on past performance data and are unable to consider the emotional responses of viewers. This makes it difficult to grasp the specific emotional impact that advertising has on viewers and to propose appropriate improvements based on that. In particular, there was a need for a method to evaluate the impact of visual and audio factors on viewer emotions, in addition to click-through rates and conversion rates.

[0312] 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.

[0313] In this invention, the server includes means for collecting performance data of past advertisements, means for collecting emotional response data, means for preprocessing the collected performance data and emotional response data, means for performing AI analysis based on the preprocessed data and emotional response data, means for extracting improvement points for the advertisement creative based on the results of the AI ​​analysis, means for generating a report including the improvement points, and means for providing the generated report to the user's connection device, thereby enabling efficient improvement of the advertisement creative based on the viewer's emotional response.

[0314] "Past advertising performance data" refers to data including performance indicators such as click rates, conversion rates, and number of impressions obtained in ad distribution.

[0315] "Preprocessing" refers to the process of preparing data suitable for analysis after collection by performing tasks such as filling in missing values, filtering, and normalizing the scale.

[0316] "Emotional response data" refers to emotional data of a user analyzed from facial expressions, tone of voice, text input, and the like while the user is viewing an advertisement.

[0317] "AI analysis" is the process of using machine learning algorithms to evaluate the impact of each element of advertising creative on performance metrics and emotions based on past advertising performance data and emotional response data.

[0318] "Ad creative" refers to the design elements of an ad, such as images, text, layout, and color usage.

[0319] "Improvement points" are specific elements or methods that are recommended to be changed or improved based on the results of AI analysis in order to improve the performance of advertising creatives.

[0320] "Report generation" is the process of visually organizing the AI ​​analysis results and points for improvement, and creating a report using charts and graphs to provide to the user.

[0321] "User's connected device" refers to the device (e.g., smartphone, tablet, PC) used by the user to provide the above-mentioned report.

[0322] A "machine learning algorithm" is an algorithm (e.g., random forest, neural network) that uses large amounts of data to train a model and make predictions or classifications from new data.

[0323] This invention is a system for efficiently improving advertising creatives, which utilizes a combination of past advertising performance data and user emotional response data. Specifically, it includes the following components:

[0324] First, the server uses an API to collect past advertising performance data from the ad distribution platform. This data includes performance indicators such as click-through rate, conversion rate, and number of impressions. Furthermore, a sentiment analysis engine is used to collect emotional responses from users' facial expressions, voice tone, text input, etc. This allows the server to also collect emotional data on users who view ads.

[0325] Next, the server preprocesses the collected ad performance data and emotional response data. Specifically, it imputes missing values, filters to maintain consistency, and normalizes the data scale. This prepares both data sets for analysis. Data preprocessing is performed using data processing libraries such as Pandas and Scikit-learn.

[0326] The server then uses machine learning algorithms to analyze the pre-processed ad performance data and emotional response data. At this stage, it creates a model that evaluates the impact of each element of the ad creative on performance indicators and emotions, using tools such as random forests and neural networks (using Tensorflow® and Keras).

[0327] Based on the analysis results, the server can identify areas for improvement in ad creatives. For example, it may determine that ads with high click-through rates and conversion rates are effective with specific colors, fonts, and layouts, and that they also evoke positive emotional responses from users.

[0328] Points for improvement are compiled into specific proposals, which are then generated by the server as a report. The report is visually organized using charts and graphs and delivered to the user's connected device (smartphone, tablet, or PC). Visualization libraries such as Matplotlib and Seaborn are used to generate the report. Reports can also be created in PDF format using ReportLab.

[0329] As a specific example, consider a case where a server collects and analyzes the following data:

[0330] Summer sale advertisement:

[0331] Click-through rate: 5%

[0332] Conversion rate: 2%

[0333] Views: 100,000

[0334] Positive emotional responses (smiles, positive comments): 70%

[0335] Fall sale ad:

[0336] Click-through rate: 1%

[0337] Conversion rate: 0.5%

[0338] Views: 80,000

[0339] Negative emotional responses (sad facial expressions, negative comments): 50%

[0340] An example of a prompt sentence is as follows:

[0341] System Input:

[0342] The system collects data for a summer sale ad with a 5% click-through rate, a 2% conversion rate, 100,000 impressions, and a 70% positive emotional response, and a fall sale ad with a 1% click-through rate, a 0.5% conversion rate, 80,000 impressions, and a 50% negative emotional response.

[0343] Inference result:

[0344] We suggest that brightening the colors of advertising images and enlarging the font size would be effective.

[0345] This will enable efficient improvement of advertising creatives based on viewers' emotional responses.

[0346] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0347] Step 1:

[0348] The server uses the API from the ad distribution platform to collect data on past ad performance and user emotional response. This data includes click-through rates, conversion rates, number of impressions, users' facial expressions, voice tone, and text input. This data is the basis for measuring the effectiveness of ads and determining users' emotional responses.

[0349] Input: API of ad distribution platform, user facial expressions, voice tone, text input

[0350] Output: Collected advertising performance data and emotional response data

[0351] Step 2:

[0352] The ad performance data and emotional response data collected by the server are preprocessed. Specifically, missing values ​​are imputed, inconsistent data is filtered, and the data scale is normalized. Data processing libraries such as Pandas and Scikit-learn are used for preprocessing. Preprocessing improves data consistency and suitability for analysis.

[0353] Input: Collected advertising performance data and emotional response data

[0354] Output: Preprocessed ad performance data and emotional response data

[0355] Step 3:

[0356] The server performs AI analysis based on the preprocessed ad performance data and emotional response data. This analysis applies machine learning algorithms such as random forests and neural networks (using TensorFlow, Keras, etc.). The AI ​​analysis evaluates the impact of ad creative elements on performance indicators and user emotional responses and creates a model.

[0357] Input: Preprocessed advertising performance data and emotional response data

[0358] Output: A model for evaluating the impact of advertising creative elements on performance metrics and emotional responses.

[0359] Step 4:

[0360] Based on the results of the AI ​​analysis, the server extracts points for improvement in the ad creative. For example, it suggests specific ways to improve elements such as color usage, font, size, and placement. At this stage, feedback is obtained to suggest effective improvement measures based on the analysis results.

[0361] Input: A model for evaluating the impact of advertising creative elements on performance indicators and emotional responses.

[0362] Output: Specific suggestions for improvement

[0363] Step 5:

[0364] The server visually organizes the identified improvement points and generates a report, which can be created in PDF or web-based dashboard format and designed to include charts and graphs using visualization libraries such as Matplotlib, Seaborn, and ReportLab.

[0365] Input: Specific suggestions for improvement

[0366] Output: Visualized report

[0367] Step 6:

[0368] The server provides the generated report to the user's connected device (smartphone, tablet, PC), allowing the user to improve their ad creative and apply it to their next ad campaign to maximize the effectiveness of their ads.

[0369] Input: Visualized report

[0370] Output: Report provided to user's connected devices

[0371] 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.

[0372] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (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.

[0373] 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.

[0374] [Second embodiment]

[0375] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0376] 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.

[0377] 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).

[0378] 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.

[0379] 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.

[0380] 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).

[0381] 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.

[0382] 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.

[0383] 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.

[0384] 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.

[0385] 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.

[0386] 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."

[0387] The present invention is a system for efficiently improving advertising creatives, and is specifically implemented as follows: The system is mainly composed of components of a server, a terminal, and a user.

[0388] Data collection and preprocessing

[0389] In the initial stage, the server collects historical ad performance data from ad serving platforms and databases. This data includes performance indicators such as click-through rates, conversion rates, and number of impressions. Once collected, the server fills in missing values, filters, and normalizes the data to prepare it for analysis. This ensures consistency and accuracy of the data.

[0390] AI analysis

[0391] The server then uses machine learning algorithms to analyze the pre-processed data. Learning models include random forests and neural networks, and evaluate the impact that specific elements of ad creative have on performance indicators. This analysis allows the system to learn the characteristics of past successful and unsuccessful ads, making it possible to predict the performance of new ads.

[0392] Extracting points for improvement

[0393] Based on the results of AI analysis, the server extracts specific points for improvement in the ad creative. For example, if it concludes that changes in color, text font, or layout will contribute to an increase in click-through rate, it presents that information as a specific improvement suggestion, providing designers and marketers with clear guidelines.

[0394] Generate and deliver reports

[0395] The server visually organizes the identified improvement points and generates a report using charts and graphs. The report is output in PDF or web-based dashboard format and is designed for easy viewing by the user. Finally, the server sends the generated report to the terminal, where the user can review it and take improvement actions.

[0396] Specific examples

[0397] For example, let's say an e-commerce site has a very successful summer sales campaign with the following performance data:

[0398] Click-through rate: 5%

[0399] Conversion rate: 2%

[0400] Views: 100,000

[0401] In contrast, here's the data for a failed fall sales campaign ad:

[0402] Click-through rate: 1%

[0403] Conversion rate: 0.5%

[0404] Views: 80,000

[0405] The server collects this data and performs preprocessing. The AI ​​analysis reveals that bright colors and highly visible fonts were effective for summer advertising. The extracted improvement points are compiled into a report as specific suggestions, such as "brighten the colors of advertising images and change the text font to a larger, more visible one."

[0406] Finally, the server provides this report to the user's device, and the user can design the next advertising campaign based on the suggested improvements, which can improve advertising efficiency. Through this process, an improvement in advertising return on investment (ROI) can be realized.

[0407] The processing flow will be explained below.

[0408] Step 1:

[0409] The server sends an API request to an ad serving platform or database to collect performance data for past ads, including performance metrics such as click-through rates, conversion rates, and impressions.

[0410] Step 2:

[0411] The server preprocesses the collected data by imputing missing values, filtering for consistency, and normalizing the data scale, making the data suitable for analysis.

[0412] Step 3:

[0413] The server then feeds the pre-processed data into machine learning algorithms, such as models like random forests or neural networks, that analyze the impact of each element of the ad creative on performance.

[0414] Step 4:

[0415] The server trains the machine learning model and checks the model's predictive accuracy through cross-validation, thereby building an accurate predictive model.

[0416] Step 5:

[0417] The server extracts specific points for improvement based on the analysis results. For example, it identifies common characteristics of ads with high click-through rates and conversion rates and summarizes them as areas for improvement. This information is presented in the form of suggestions for color, font, layout, etc.

[0418] Step 6:

[0419] The server visually organizes the identified improvement points and generates a report using charts and graphs. The report is created in PDF or web-based dashboard format, with an emphasis on ease of viewing and understanding.

[0420] Step 7:

[0421] The server sends the generated report to the user's terminal, which allows the user to quickly take concrete action regarding the creation of new advertisements.

[0422] Step 8:

[0423] We review the reports received by the user and design and produce new ad creatives based on the suggested improvements. Based on the suggestions, we make specific improvements such as changes to color usage, layout, and text fonts.

[0424] Step 9:

[0425] The user then redistributes the improved ad, collecting new performance data. This data is then sent back to the server for analysis in the next cycle. By repeating this process, ad performance continues to improve.

[0426] Example 1

[0427] 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."

[0428] Improving the performance of advertising creatives is important in marketing, but conventional methods have made it difficult to quickly and accurately identify effective areas for improvement. To solve this problem, there is a need to develop a system that can efficiently and accurately extract and suggest areas for improvement using past advertising performance data.

[0429] 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.

[0430] In this invention, the server includes a means for collecting performance data of past advertisements, a means for preprocessing the collected data, and a means for performing AI analysis based on the preprocessed data, which makes it possible to evaluate the impact of elements of advertising creatives on performance indicators and quickly and accurately extract areas for improvement.

[0431] "Past advertising performance data" refers to data that includes advertising performance indicators such as click rates, conversion rates, and number of impressions.

[0432] "Means of collection" refers to the means for obtaining past advertising performance data from advertising distribution platforms and databases.

[0433] The "preprocessing means" refers to means for imputing missing values ​​in the collected data, filtering the data, and normalizing the data.

[0434] "Means for conducting AI analysis" means means for applying machine learning algorithms using pre-processed data to evaluate the impact of elements of advertising creative on performance metrics.

[0435] "Means for extracting points for improvement in advertising creative" refers to a means for identifying specific changes to improve the effectiveness of advertising based on the results of AI analysis.

[0436] The "means for generating a report" refers to a means for visually organizing the extracted improvement points and creating a report using charts and graphs.

[0437] The "means for providing to the user's terminal" refers to the means for providing the generated report in a form that can be viewed by the user, such as a PDF format or a web-based dashboard.

[0438] A "data analysis algorithm" is a mathematical model or computational method for finding patterns and trends in data and making predictions.

[0439] "Data reconciliation" is the process of converting data collected from different data sources into a unified format and making it consistent.

[0440] "Data standardization" is the process of scaling all numerical data to a certain range, typically normalizing it to the range 0 to 1.

[0441] This invention is a system aimed at efficiently improving advertising creatives. The system is mainly composed of components such as a server, a terminal, and a user.

[0442] Data collection and preprocessing

[0443] The server collects past advertising performance data from ad distribution platforms and databases. This data includes performance indicators such as click-through rates, conversion rates, and number of impressions. A specific example of this data collection is using the Google Ads API. The API is used to obtain advertising performance data for a specified period and store it in a database.

[0444] The server performs missing value imputation, filtering, and normalization on the collected data. This preprocessing is performed using the Python Pandas library. Specifically, the following operations are performed:

[0445] Use the mean or median to impute missing values.

[0446] To remove unnecessary data, filtering is performed according to specific conditions.

[0447] Scale all numeric data to the range 0 to 1.

[0448] AI analysis

[0449] The server runs machine learning algorithms on the preprocessed data, such as random forests and neural networks. Specifically, it uses the Scikit-learn library to train a random forest model to predict ad performance.

[0450] The server performs the following operations:

[0451] Using the Scikit-learn library, training data is input into the model and learning is performed.

[0452] Use the trained model to predict performance on new advertising data.

[0453] Extracting points for improvement

[0454] Based on the results of the AI ​​analysis, the server extracts specific points for improvement in the ad creative. From the analysis results, it identifies the factors that have the greatest impact on performance and forms specific improvement proposals based on those.

[0455] Specific suggestions include using bright colors and highly visible fonts. This information is provided as useful feedback to designers and marketers.

[0456] Generate and deliver reports

[0457] The server visually organizes the identified improvement points and generates a report using charts and graphs. This report is output in PDF format and a web-based dashboard format. The Matplotlib library is used to generate various graphs and convert them to PDF.

[0458] The server finally delivers this report to the user's device, either by emailing the generated PDF file or uploading it to a web dashboard where the user can log in and view it.

[0459] Specific examples

[0460] For example, let's say an e-commerce site runs a successful summer sales campaign with the following performance data:

[0461] Click-through rate: 5%

[0462] Conversion rate: 2%

[0463] Views: 100,000

[0464] In contrast, here's the data for a failed fall sales campaign ad:

[0465] Click-through rate: 1%

[0466] Conversion rate: 0.5%

[0467] Views: 80,000

[0468] The server collects this data and performs preprocessing. The AI ​​analysis reveals that bright colors and highly visible fonts were effective for summer advertising. The extracted improvement points are compiled into a report as specific suggestions, such as "brighten the colors of advertising images and change the text font to a larger, more visible one."

[0469] Prompt Sentence Examples

[0470] Users are prompted to take specific actions, such as, "Use bright colors and large, highly visible text fonts in your next ad campaign images to increase click-through rates and conversion rates."

[0471] This series of systems makes it possible to improve advertising return on investment (ROI).

[0472] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0473] Step 1: Data collection

[0474] The first thing the server does is collect past advertising performance data from ad distribution platforms and databases. Specifically, it uses the Google Ads API to obtain data such as click-through rates, conversion rates, and number of impressions for a specified period. The input data is the unprocessed raw data obtained from the API, and the output is past advertising performance data stored in the server's database.

[0475] Specific behavior:

[0476] Access the Google Ads API using an authentication token to retrieve the required data.

[0477] The acquired data is saved in a CSV file or database.

[0478] Step 2: Data Preprocessing

[0479] The server then preprocesses the collected data, which includes missing value imputation, data filtering, and normalization. The input to the preprocessing is the raw data collected, and the output is a consistent and accurate dataset that can be analyzed.

[0480] Specific behavior:

[0481] Create a data frame using Python's Pandas library.

[0482] Impute missing values ​​with the mean or median.

[0483] Filter data based on specific criteria to remove unwanted data.

[0484] Scales numeric data to the range 0 to 1.

[0485] Step 3: AI analysis

[0486] The server runs a machine learning algorithm on the preprocessed data. This step uses models like random forests or neural networks. The input is the preprocessed dataset, and the output is an analysis showing the impact of each element of the ad creative on performance metrics.

[0487] Specific behavior:

[0488] Train a random forest model using the Scikit-learn library.

[0489] Use the trained model to predict performance on new advertising data.

[0490] Step 4: Identifying areas for improvement

[0491] The server extracts specific points for improvement in the ad creative based on the results of the AI ​​analysis. The input is the AI ​​analysis results, and the output is specific improvement suggestions.

[0492] Specific behavior:

[0493] The factors with the greatest impact are extracted from the analysis results.

[0494] Create specific improvement suggestions for the use of colors in advertising images and text fonts.

[0495] Step 5: Generate reports

[0496] The server visually organizes the extracted improvement points and generates a report using charts and graphs. The input is the improvement proposals, and the output is a report in PDF format or a web dashboard format.

[0497] Specific behavior:

[0498] Generate various graphs using the Matplotlib library.

[0499] Convert the generated graphs to PDF and compile a report.

[0500] Step 6: Providing the report

[0501] The server provides the generated report to the user's device. This step involves emailing the PDF file or uploading it to a web dashboard. The input is the generated report and the output is the report provided to the user.

[0502] Specific behavior:

[0503] Generate a PDF report and send it to the user via email.

[0504] Upload reports to a web dashboard so users can log in and view them.

[0505] (Application example 1)

[0506] 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."

[0507] Conventional methods for improving advertising creatives require time and effort to collect and analyze past performance data, making it difficult to propose effective improvements in real time. Furthermore, since specific improvement proposals based on the expected effectiveness of advertising creatives cannot be obtained, there are limited ways to maximize advertising performance.

[0508] 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.

[0509] In this invention, the server includes means for collecting performance data of past advertisements, means for preprocessing the collected data, means for performing AI analysis based on the preprocessed data, means for extracting improvement points for the advertisement creative based on the results of the AI ​​analysis, means for generating a report including the improvement points, means for providing the generated report to a user's terminal, means for providing improvement suggestions for the advertisement creative in real time, and means for generating responsive prompt sentences. This enables efficient creative improvement in real time based on the performance data of advertisements, thereby maximizing advertisement performance.

[0510] "Past advertising performance data" refers to data including performance indicators related to past advertising distribution, such as click rates, conversion rates, and number of impressions of advertisements.

[0511] "Data preprocessing" is the process of completing the collected data by imputing missing values, filtering, and normalizing it, preparing it for analysis.

[0512] "AI analysis" is the process of analyzing data using machine learning algorithms to assess the impact that specific elements of ad creative have on performance metrics.

[0513] "Improvement points" are specific suggested changes to improve the performance of ad creatives, extracted based on the results of AI analysis.

[0514] "Report generation" is the process of visually organizing improvement points and outputting them in a format that is easy for users to understand using charts and graphs.

[0515] "User device" refers to an electronic device owned by a user, such as a smartphone, tablet, or PC, to which a report is provided.

[0516] "Means for providing real-time improvement suggestions for advertising creatives" is a function for quickly providing users with improvement suggestions for advertising creatives that are generated based on the results of AI analysis.

[0517] The "means for generating prompt sentences" is a function that automatically creates sentences to make suggestions or ask questions to the user in order to support the improvement of advertising creatives.

[0518] This invention is a system for realizing efficient improvement of advertising creatives, and is composed of the following components: The system includes the following elements: a server, a terminal, and a user.

[0519] Data collection and preprocessing

[0520] In the initial stage, the server collects historical ad performance data from ad serving platforms and databases. This data includes performance indicators such as click-through rates, conversion rates, and number of impressions. Once collected, the server fills in missing values, filters, and normalizes the data to prepare it for analysis. This ensures consistency and accuracy of the data.

[0521] AI analysis

[0522] After data preprocessing is complete, the server performs analysis using machine learning algorithms. Learning models used include random forests and neural networks. This allows the impact of specific elements of ad creative on performance indicators to be evaluated. AI analysis learns the characteristics of past successful and unsuccessful ads, allowing for prediction of the performance of new ads.

[0523] Extracting points for improvement

[0524] Based on the results of the AI ​​analysis, the server extracts specific points for improvement in the ad creative. For example, if it concludes that changes in color, text font, or layout will contribute to an increase in click-through rate, it will present that information as a specific improvement suggestion.

[0525] Generate and deliver reports

[0526] The server visually organizes the extracted improvement points and generates a report using charts and graphs. The report is output in PDF or web-based dashboard format and is designed to be easy for users to view. Finally, the server sends the generated report to the terminal, where the user can review it and take improvement actions.

[0527] Real-time improvement suggestions for advertising creatives

[0528] Based on AI analysis, the server generates real-time suggestions for improving ad creatives and provides them to users' devices. When users create new ads, they can make specific changes while receiving immediate feedback from the server.

[0529] Generate prompt statement

[0530] Furthermore, the server can generate prompt sentences to assist the user in creating an advertisement. The prompt sentences are sentences that suggest improvements to the advertisement creative and serve to prompt the user to take specific action to improve the advertisement.

[0531] Specific examples

[0532] For example, an e-commerce site may have a very successful summer sales campaign with the following performance data:

[0533] Click-through rate: 5%

[0534] Conversion rate: 2%

[0535] Views: 100,000

[0536] In contrast, here's the data for a failed fall sales campaign ad:

[0537] Click-through rate: 1%

[0538] Conversion rate: 0.5%

[0539] Views: 80,000

[0540] The server collects this data and performs preprocessing. The AI ​​analysis reveals that bright colors and highly visible fonts were effective for summer advertising. The extracted improvement points are compiled into a report as specific suggestions, such as "brighten the colors of advertising images and change the text font to a larger, more visible one." Finally, the server provides this report to the user's device, and the user can design the next advertising campaign based on the suggested improvements, which is expected to improve advertising efficiency.

[0541] Also, when a user creates a new ad creative, the server provides a prompt like this:

[0542] "We recommend using brighter colors and more legible fonts in your new ads. This change may increase your click-through rate."

[0543] This allows the system to help improve effective advertising creative in real time.

[0544] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0545] Step 1:

[0546] The server collects performance data of past ads from ad distribution platforms and databases. The collected data includes click-through rates, conversion rates, number of impressions, etc. The input is the performance data of the ads, and the output is the collected raw data.

[0547] Step 2:

[0548] The server preprocesses the collected data. Specifically, it imputes missing values ​​in the data, filters to remove noise, and scales and normalizes the data. The input is the raw data collected in step 1, and the output is the preprocessed data.

[0549] Step 3:

[0550] The server uses the preprocessed data to perform AI analysis, using machine learning models such as random forests and neural networks to evaluate the impact of specific elements of the ad creative on performance metrics. The input is the preprocessed data, and the output is the analysis results.

[0551] Step 4:

[0552] The server extracts specific points for improvement in the ad creative based on the results of the AI ​​analysis. For example, it generates specific suggestions, such as how changing the color usage or text font will contribute to improving click-through rates. The input is the analysis results obtained in step 3, and the output is the points for improvement.

[0553] Step 5:

[0554] The server generates a visually organized report using the extracted improvement points. This report is output in a user-friendly format using charts and graphs. The input is the improvement points extracted in step 4, and the output is the report.

[0555] Step 6:

[0556] The server provides the generated report to the user's device, allowing the user to review the report and take specific improvement actions. The input is the report generated in step 5, and the output is the report delivered to the user's device.

[0557] Step 7:

[0558] The server provides real-time suggestions for improving ad creatives. Users can receive immediate feedback when creating new ad creatives. The input is the initial data for new ad creatives and analysis results based on past performance data, and the output is real-time suggestions for improvement.

[0559] Step 8:

[0560] The server generates a prompt to assist the user in creating an ad. The prompt serves to encourage the user to take specific action to improve the ad. The input is the improvement proposal obtained in step 7, and the output is the prompt. A specific example would be, "We recommend that you use brighter colors in your new ad and change the font to one that is more visible. This change may improve your click-through rate."

[0561] 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.

[0562] This invention is a system aimed at efficiently improving advertising creatives, utilizing both past advertising performance data and user emotional data. The system is primarily comprised of a server, a terminal, and a user. Here, we will explain in detail how this system is implemented.

[0563] Data collection and preprocessing

[0564] At the initial stage, the server sends an API request to an ad serving platform or database to collect performance data of past ads. The collected data includes performance indicators such as click-through rate, conversion rate, and number of impressions. In addition, the server uses an emotion engine to obtain emotional data from the user's facial expressions, voice tone, text input, etc. This also collects the emotions of the user viewing the ad.

[0565] Data Preprocessing

[0566] The server preprocesses the advertising performance data and sentiment data collected. Specifically, it fills in missing values ​​in the data and performs filtering to maintain consistency. It also normalizes the data scale and prepares it for analysis. This makes both data suitable for analysis.

[0567] AI analysis

[0568] The server uses machine learning algorithms to analyze the preprocessed advertising performance data and emotion data. The learning models include random forests and neural networks, and evaluate the impact of each element of the advertising creative on performance indicators and user emotions. This analysis learns the characteristics of past successful and unsuccessful ads, making it possible to predict the performance and evaluate emotions of new ads.

[0569] Extracting points for improvement

[0570] Based on the results of the AI ​​analysis, the server extracts specific areas for improvement in the ad creative. For example, it identifies common characteristics of ads with high click-through rates and conversion rates, as well as the corresponding positive emotional responses, and summarizes these as areas for improvement. This information is presented in the form of suggestions for color usage, fonts, layout, etc.

[0571] Generate and deliver reports

[0572] The server visually organizes the extracted improvement points and generates a report using charts and graphs. The report is created in PDF or web-based dashboard format, emphasizing ease of viewing and understanding. It also includes the results of sentiment data analysis, showing how advertising affects users' emotions. Finally, the server provides the generated report to the user's device.

[0573] Specific examples

[0574] For example, let's say an e-commerce site's summer sales campaign was extremely successful, resulting in the following performance and sentiment data:

[0575] Click-through rate: 5%

[0576] Conversion rate: 2%

[0577] Views: 100,000

[0578] Positive emotional responses (smiles, positive comments): 70%

[0579] In contrast, here's the data for a failed fall sales campaign ad:

[0580] Click-through rate: 1%

[0581] Conversion rate: 0.5%

[0582] Views: 80,000

[0583] Negative emotional responses (sad facial expressions, negative comments): 50%

[0584] The server collects this data and performs preprocessing. The AI ​​analysis results reveal that bright colors and highly visible fonts are effective in summer advertisements and also evoke positive emotional responses from users. The extracted improvement points are compiled into a report as specific suggestions, such as "brighten the colors of the advertisement images and change the text font to a larger, more visible one."

[0585] Finally, the server provides this report to the user's device, and the next advertising campaign is designed based on the user's suggested improvements, which is expected to improve advertising efficiency. Through this series of processes, advertising return on investment (ROI) and advertising that takes user emotions into consideration are realized.

[0586] The processing flow will be explained below.

[0587] Step 1:

[0588] The server sends an API request to an ad serving platform or database to collect performance data of past ads. The collected data includes performance indicators such as click-through rate, conversion rate, and number of impressions. At the same time, an emotion engine is run to obtain emotional data from the user's facial expressions, tone of voice, text input, etc.

[0589] Step 2:

[0590] The server preprocesses the advertising performance data and sentiment data collected. First, if any missing values ​​exist, they are imputed appropriately. Next, the data is filtered to ensure consistency. Finally, the scale of the input data is normalized and converted into a format suitable for machine learning algorithms.

[0591] Step 3:

[0592] The server then feeds the pre-processed data into machine learning algorithms, such as random forests and neural networks, to evaluate the impact of each element of the ad creative on performance metrics and user sentiment.

[0593] Step 4:

[0594] The server trains the machine learning model and performs cross-validation, checking the accuracy of the model based on the results of the cross-validation and optimizing the model as needed, thereby building a more accurate predictive model.

[0595] Step 5:

[0596] Based on the analysis results, the server extracts specific points for improvement in the ad creative. For example, it identifies common characteristics of ads with high click-through rates and conversion rates, as well as elements that evoke positive emotional responses. This information is presented as improvement suggestions in the form of specific things like color usage, fonts, and layout.

[0597] Step 6:

[0598] The server visually organizes the identified improvement points and generates a report using charts and graphs in PDF or web-based dashboard format, which also includes the results of sentiment data analysis to make it easier to understand the overall picture.

[0599] Step 7:

[0600] The server sends the generated report to the user's device, allowing the user to review the report and use it to design and improve new ad creatives.

[0601] Step 8:

[0602] We carefully review the reports received by users and design and produce new ad creatives based on the suggested improvements. Based on the suggestions, we make specific improvements such as changes to color usage, layout, and text fonts.

[0603] Step 9:

[0604] The user then redistributes the improved ad, collecting new performance data. This new data is then sent back to the server for analysis in the next cycle. This process is repeated, resulting in continuous improvements in ad performance and user emotional response.

[0605] Example 2

[0606] 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."

[0607] Conventional advertising improvement systems have been unable to effectively improve advertising performance because they have been unable to fully integrate past advertising performance data and user emotion data, and their specific proposals for improving advertising creatives are abstract. Furthermore, it has been difficult to efficiently collect and analyze user emotion data and perform emotion evaluation based on advertising performance.

[0608] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting past advertising performance data, means for pre-processing the collected data, means for collecting user emotion data, means for performing AI analysis based on the pre-processed data, means for extracting improvement points for advertising creative based on the results of the AI ​​analysis, means for generating a report including the improvement points, and means for providing the generated report to the user's terminal. This makes it possible to comprehensively analyze past advertising performance data and user emotion data and effectively improve advertising performance through specific improvement suggestions.

[0609] "Advertising performance data" refers to performance indicators such as click rates, conversion rates, and number of impressions that indicate the effectiveness of past advertising campaigns.

[0610] "Preprocessing" refers to the process of filling in missing values ​​in the data, filtering to maintain consistency, and normalizing the data to prepare it for analysis.

[0611] "User emotion data" refers to data that indicates the user's emotional state, obtained from the user's facial expression, voice tone, text input, and the like.

[0612] "AI analytics" is the process of using machine learning algorithms to analyze pre-processed data and evaluate the impact of each element of advertising creative on performance metrics and user sentiment.

[0613] "Points for improving advertising creative" refers to specific changes and improvements that are needed to improve advertising performance based on the results of AI analysis.

[0614] A "report" is a report that visually organizes the analysis results and extracted improvement points and displays them using charts, graphs, etc.

[0615] A "user's terminal" is a device such as a computer, smartphone, or tablet that a user uses to receive reports provided by the server.

[0616] This invention is a system aimed at efficiently improving advertising creatives, and provides a means for integrating and analyzing past advertising performance data and user emotion data. The system is primarily comprised of components: a server, a terminal, and a user. Here, we will explain in detail how this system is implemented.

[0617] Data collection

[0618] The server first sends an API request to an ad serving platform or database to collect past ad performance data. Specific APIs typically used are the Google Ads API and Facebook Ads API. This data includes click-through rates, conversion rates, and number of impressions.

[0619] The server also collects the user's emotional data using an emotion engine, such as Microsoft Azure Emotion API or IBM Watson Tone Analyzer, which analyzes and obtains the user's emotional state from their facial expressions, voice tone, and text input. Emotional data is collected when the device captures the user's video and audio and sends them to the server.

[0620] Data Preprocessing

[0621] The server preprocesses the advertising performance data and sentiment data collected by the server. The Python pandas library is used to impute missing values, and data filtering and normalization are also performed. For example, this includes outlier detection and normalization of the data scale using a scaling library.

[0622] AI analysis

[0623] The server then analyzes the pre-processed data using machine learning algorithms, specifically models such as random forests and neural networks. The analysis uses the scikit-learn library to train and evaluate the models. The AI ​​analysis evaluates the impact of each element of the ad creative on performance metrics and user sentiment.

[0624] Extracting points for improvement

[0625] The server extracts specific areas for improvement in the ad creative based on the results of the AI ​​analysis. It identifies common characteristics of ads with high click-through rates and conversion rates and summarizes the areas for improvement. For example, it generates suggestions for "bright colors" and "highly visible fonts."

[0626] Generate and deliver reports

[0627] The server visually organizes the extracted improvement points and generates a report using charts and graphs. Matplotlib and Plotly are used to generate the report, which is automatically created using a PDF generator. Finally, the server provides the generated report to the user's device. The report is delivered as a download link or email attachment, and the user can use it to design their next advertising campaign.

[0628] Specific examples

[0629] For example, suppose an e-commerce site runs a summer sale campaign ad with a 5% click-through rate, a 2% conversion rate, 100,000 impressions, and a 70% positive emotional response. In contrast, if a fall sale campaign ad has a 1% click-through rate, a 0.5% conversion rate, 80,000 impressions, and a 50% negative emotional response, the server collects and preprocesses this data and performs AI analysis. The analysis results suggest that bright colors and highly visible fonts are effective for summer ads.

[0630] The extracted improvement points are compiled into a report with specific suggestions, such as "Use brighter colors in the ad images and change the text font to a larger, more visible one." Finally, the server provides the generated report to the user's device, allowing the user to incorporate the improvements into their next ad campaign.

[0631] Prompt Sentence Examples

[0632] Please suggest areas for improvement in the ad creative based on the advertising data below.

[0633] (Successful summer sale campaign advertising data)

[0634] Click-through rate: 5%

[0635] Conversion rate: 2%

[0636] Views: 100,000

[0637] Positive emotional response: 70%

[0638] (Failed Fall Sale Campaign Advertising Data)

[0639] Click-through rate: 1%

[0640] Conversion rate: 0.5%

[0641] Views: 80,000

[0642] Negative emotional reactions: 50%

[0643] suggestion:

[0644] Use bright colors and highly visible fonts, just like in summer ads, to increase click-through rates and conversion rates.

[0645] With the above configuration, this system improves advertising performance and proposes advertising creatives that take user emotions into consideration.

[0646] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0647] Step 1: Collect advertising performance data

[0648] The server sends API requests to ad distribution platforms and databases to collect past ad performance data. Specifically, it uses the Google Ads API and Facebook Ads API. The input is API authentication information and the query required to obtain data, and the output is data on ad performance indicators such as click-through rate (CTR), conversion rate (CVR), and number of impressions. The server performs API authentication using the client ID and secret, and obtains data from the specified endpoint.

[0649] Step 2: Collecting emotion data

[0650] The server runs an emotion engine and collects emotion data from the user's facial expressions, voice tone, and text input. The inputs are camera footage, audio clips, and text data, and the emotional state is analyzed based on this. Specifically, the Microsoft Azure Emotion API and IBM Watson Tone Analyzer are used. The device captures the user's video and audio and sends them to the server to obtain emotion data. The output is data indicating the user's emotional state (e.g., smile rate, negative comment rate).

[0651] Step 3: Preprocessing the data

[0652] The advertising performance data and sentiment data collected by the server are preprocessed. The input is the raw data obtained in steps 1 and 2. Specific preprocessing includes missing value imputation, filtering, and normalization. The Python pandas library is used to impute missing values ​​with the mean or median, and outliers are filtered. Furthermore, a scaling library is used to normalize the data. The output is data that is ready for analysis.

[0653] Step 4: AI analysis

[0654] The server applies a machine learning algorithm to the preprocessed data and performs analysis. The input is the preprocessed advertising performance data and emotion data. Specifically, a random forest or neural network model is used, and the model is trained and evaluated using the scikit-learn library. The output is an evaluation of the impact of each element of the advertising creative on performance indicators and user emotions. The server randomly shuffles the training data and divides it into a training set and a test set. The algorithm is applied to train the model, and accuracy is verified on the test set.

[0655] Step 5: Identifying areas for improvement

[0656] The server extracts specific improvement points for the ad creative based on the results of the AI ​​analysis. The input is the data results of the AI ​​analysis. The server analyzes the importance of the features and displays a ranking of the elements that have the greatest impact on performance. The improvement points are written out as specific suggestions using natural language generation (NLG) technology. The output is specific changes and improvements (e.g., bright colors, highly visible fonts) as improvement suggestions.

[0657] Step 6: Generate reports

[0658] The server visually organizes the improvement points extracted and generates a report using charts and graphs. The input is improvement point data. Matplotlib and Plotly are used to generate the report. A data visualization library is used to graphically represent each indicator, and a PDF generator is used to automatically generate the report. The output is a report in PDF format or a web-based dashboard format.

[0659] Step 7: Providing the report

[0660] The server provides the generated report to the user's device. The input is the generated report data. The server configures the SMTP settings to send the report to the user's email address, and includes links and attachments in the email body. The output is a report provided to the user's device, which the user can use to design their next advertising campaign.

[0661] (Application example 2)

[0662] 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."

[0663] Conventional systems for improving advertising creatives rely solely on past performance data and are unable to consider the emotional responses of viewers. This makes it difficult to grasp the specific emotional impact that advertising has on viewers and to propose appropriate improvements based on that. In particular, there was a need for a method to evaluate the impact of visual and audio factors on viewer emotions, in addition to click-through rates and conversion rates.

[0664] 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.

[0665] In this invention, the server includes means for collecting performance data of past advertisements, means for collecting emotional response data, means for preprocessing the collected performance data and emotional response data, means for performing AI analysis based on the preprocessed data and emotional response data, means for extracting improvement points for the advertisement creative based on the results of the AI ​​analysis, means for generating a report including the improvement points, and means for providing the generated report to the user's connection device, thereby enabling efficient improvement of the advertisement creative based on the viewer's emotional response.

[0666] "Past advertising performance data" refers to data including performance indicators such as click rates, conversion rates, and number of impressions obtained in ad distribution.

[0667] "Preprocessing" refers to the process of preparing data suitable for analysis after collection by performing tasks such as filling in missing values, filtering, and normalizing the scale.

[0668] "Emotional response data" refers to emotional data of a user analyzed from facial expressions, tone of voice, text input, and the like while the user is viewing an advertisement.

[0669] "AI analysis" is the process of using machine learning algorithms to evaluate the impact of each element of advertising creative on performance metrics and emotions based on past advertising performance data and emotional response data.

[0670] "Ad creative" refers to the design elements of an ad, such as images, text, layout, and color usage.

[0671] "Improvement points" are specific elements or methods that are recommended to be changed or improved based on the results of AI analysis in order to improve the performance of advertising creatives.

[0672] "Report generation" is the process of visually organizing the AI ​​analysis results and points for improvement, and creating a report using charts and graphs to provide to the user.

[0673] "User's connected device" refers to the device (e.g., smartphone, tablet, PC) used by the user to provide the above-mentioned report.

[0674] A "machine learning algorithm" is an algorithm (e.g., random forest, neural network) that uses large amounts of data to train a model and make predictions or classifications from new data.

[0675] This invention is a system for efficiently improving advertising creatives, which utilizes a combination of past advertising performance data and user emotional response data. Specifically, it includes the following components:

[0676] First, the server uses an API to collect past advertising performance data from the ad distribution platform. This data includes performance indicators such as click-through rate, conversion rate, and number of impressions. Furthermore, a sentiment analysis engine is used to collect emotional responses from users' facial expressions, voice tone, text input, etc. This allows the server to also collect emotional data on users who view ads.

[0677] Next, the server preprocesses the collected ad performance data and emotional response data. Specifically, it imputes missing values, filters to maintain consistency, and normalizes the data scale. This prepares both data sets for analysis. Data preprocessing is performed using data processing libraries such as Pandas and Scikit-learn.

[0678] The server then uses machine learning algorithms to analyze the pre-processed ad performance data and emotional response data. At this stage, it uses tools such as random forests and neural networks (using TensorFlow and Keras) to create models that evaluate the impact of each element of the ad creative on performance indicators and emotions.

[0679] Based on the analysis results, the server can identify areas for improvement in ad creatives. For example, it may determine that ads with high click-through rates and conversion rates are effective with specific colors, fonts, and layouts, and that they also evoke positive emotional responses from users.

[0680] Points for improvement are compiled into specific proposals, which are then generated by the server as a report. The report is visually organized using charts and graphs and delivered to the user's connected device (smartphone, tablet, or PC). Visualization libraries such as Matplotlib and Seaborn are used to generate the report. Reports can also be created in PDF format using ReportLab.

[0681] As a specific example, consider a case where a server collects and analyzes the following data:

[0682] Summer sale advertisement:

[0683] Click-through rate: 5%

[0684] Conversion rate: 2%

[0685] Views: 100,000

[0686] Positive emotional responses (smiles, positive comments): 70%

[0687] Fall sale ad:

[0688] Click-through rate: 1%

[0689] Conversion rate: 0.5%

[0690] Views: 80,000

[0691] Negative emotional responses (sad facial expressions, negative comments): 50%

[0692] An example of a prompt sentence is as follows:

[0693] System Input:

[0694] The system collects data for a summer sale ad with a 5% click-through rate, a 2% conversion rate, 100,000 impressions, and a 70% positive emotional response, and a fall sale ad with a 1% click-through rate, a 0.5% conversion rate, 80,000 impressions, and a 50% negative emotional response.

[0695] Inference result:

[0696] We suggest that brightening the colors of advertising images and enlarging the font size would be effective.

[0697] This will enable efficient improvement of advertising creatives based on viewers' emotional responses.

[0698] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0699] Step 1:

[0700] The server uses the API from the ad distribution platform to collect data on past ad performance and user emotional response. This data includes click-through rates, conversion rates, number of impressions, users' facial expressions, voice tone, and text input. This data is the basis for measuring the effectiveness of ads and determining users' emotional responses.

[0701] Input: API of ad distribution platform, user facial expressions, voice tone, text input

[0702] Output: Collected advertising performance data and emotional response data

[0703] Step 2:

[0704] The ad performance data and emotional response data collected by the server are preprocessed. Specifically, missing values ​​are imputed, inconsistent data is filtered, and the data scale is normalized. Data processing libraries such as Pandas and Scikit-learn are used for preprocessing. Preprocessing improves data consistency and suitability for analysis.

[0705] Input: Collected advertising performance data and emotional response data

[0706] Output: Preprocessed ad performance data and emotional response data

[0707] Step 3:

[0708] The server performs AI analysis based on the preprocessed ad performance data and emotional response data. This analysis applies machine learning algorithms such as random forests and neural networks (using TensorFlow, Keras, etc.). The AI ​​analysis evaluates the impact of ad creative elements on performance indicators and user emotional responses and creates a model.

[0709] Input: Preprocessed advertising performance data and emotional response data

[0710] Output: A model for evaluating the impact of advertising creative elements on performance metrics and emotional responses.

[0711] Step 4:

[0712] Based on the results of the AI ​​analysis, the server extracts points for improvement in the ad creative. For example, it suggests specific ways to improve elements such as color usage, font, size, and placement. At this stage, feedback is obtained to suggest effective improvement measures based on the analysis results.

[0713] Input: A model for evaluating the impact of advertising creative elements on performance indicators and emotional responses.

[0714] Output: Specific suggestions for improvement

[0715] Step 5:

[0716] The server visually organizes the identified improvement points and generates a report, which can be created in PDF or web-based dashboard format and designed to include charts and graphs using visualization libraries such as Matplotlib, Seaborn, and ReportLab.

[0717] Input: Specific suggestions for improvement

[0718] Output: Visualized report

[0719] Step 6:

[0720] The server provides the generated report to the user's connected device (smartphone, tablet, PC), allowing the user to improve their ad creative and apply it to their next ad campaign to maximize the effectiveness of their ads.

[0721] Input: Visualized report

[0722] Output: Report provided to user's connected devices

[0723] 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.

[0724] 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.

[0725] 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.

[0726] [Third embodiment]

[0727] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0728] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0729] 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).

[0730] 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.

[0731] 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.

[0732] 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).

[0733] 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.

[0734] 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.

[0735] 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.

[0736] 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.

[0737] 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.

[0738] 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."

[0739] The present invention is a system for efficiently improving advertising creatives, and is specifically implemented as follows: The system is mainly composed of components of a server, a terminal, and a user.

[0740] Data collection and preprocessing

[0741] In the initial stage, the server collects historical ad performance data from ad serving platforms and databases. This data includes performance indicators such as click-through rates, conversion rates, and number of impressions. Once collected, the server fills in missing values, filters, and normalizes the data to prepare it for analysis. This ensures consistency and accuracy of the data.

[0742] AI analysis

[0743] The server then uses machine learning algorithms to analyze the pre-processed data. Learning models include random forests and neural networks, and evaluate the impact that specific elements of ad creative have on performance indicators. This analysis allows the system to learn the characteristics of past successful and unsuccessful ads, making it possible to predict the performance of new ads.

[0744] Extracting points for improvement

[0745] Based on the results of AI analysis, the server extracts specific points for improvement in the ad creative. For example, if it concludes that changes in color, text font, or layout will contribute to an increase in click-through rate, it presents that information as a specific improvement suggestion, providing designers and marketers with clear guidelines.

[0746] Generate and deliver reports

[0747] The server visually organizes the identified improvement points and generates a report using charts and graphs. The report is output in PDF or web-based dashboard format and is designed for easy viewing by the user. Finally, the server sends the generated report to the terminal, where the user can review it and take improvement actions.

[0748] Specific examples

[0749] For example, let's say an e-commerce site has a very successful summer sales campaign with the following performance data:

[0750] Click-through rate: 5%

[0751] Conversion rate: 2%

[0752] Views: 100,000

[0753] In contrast, here's the data for a failed fall sales campaign ad:

[0754] Click-through rate: 1%

[0755] Conversion rate: 0.5%

[0756] Views: 80,000

[0757] The server collects this data and performs preprocessing. The AI ​​analysis reveals that bright colors and highly visible fonts were effective for summer advertising. The extracted improvement points are compiled into a report as specific suggestions, such as "brighten the colors of advertising images and change the text font to a larger, more visible one."

[0758] Finally, the server provides this report to the user's device, and the user can design the next advertising campaign based on the suggested improvements, which can improve advertising efficiency. Through this process, an improvement in advertising return on investment (ROI) can be realized.

[0759] The processing flow will be explained below.

[0760] Step 1:

[0761] The server sends an API request to an ad serving platform or database to collect performance data for past ads, including performance metrics such as click-through rates, conversion rates, and impressions.

[0762] Step 2:

[0763] The server preprocesses the collected data by imputing missing values, filtering for consistency, and normalizing the data scale, making the data suitable for analysis.

[0764] Step 3:

[0765] The server then feeds the pre-processed data into machine learning algorithms, such as models like random forests or neural networks, that analyze the impact of each element of the ad creative on performance.

[0766] Step 4:

[0767] The server trains the machine learning model and checks the model's predictive accuracy through cross-validation, thereby building an accurate predictive model.

[0768] Step 5:

[0769] The server extracts specific points for improvement based on the analysis results. For example, it identifies common characteristics of ads with high click-through rates and conversion rates and summarizes them as areas for improvement. This information is presented in the form of suggestions for color, font, layout, etc.

[0770] Step 6:

[0771] The server visually organizes the identified improvement points and generates a report using charts and graphs. The report is created in PDF or web-based dashboard format, with an emphasis on ease of viewing and understanding.

[0772] Step 7:

[0773] The server sends the generated report to the user's terminal, which allows the user to quickly take concrete action regarding the creation of new advertisements.

[0774] Step 8:

[0775] We review the reports received by the user and design and produce new ad creatives based on the suggested improvements. Based on the suggestions, we make specific improvements such as changes to color usage, layout, and text fonts.

[0776] Step 9:

[0777] The user then redistributes the improved ad, collecting new performance data. This data is then sent back to the server for analysis in the next cycle. By repeating this process, ad performance continues to improve.

[0778] Example 1

[0779] 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."

[0780] Improving the performance of advertising creatives is important in marketing, but conventional methods have made it difficult to quickly and accurately identify effective areas for improvement. To solve this problem, there is a need to develop a system that can efficiently and accurately extract and suggest areas for improvement using past advertising performance data.

[0781] 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.

[0782] In this invention, the server includes a means for collecting performance data of past advertisements, a means for preprocessing the collected data, and a means for performing AI analysis based on the preprocessed data, which makes it possible to evaluate the impact of elements of advertising creatives on performance indicators and quickly and accurately extract areas for improvement.

[0783] "Past advertising performance data" refers to data that includes advertising performance indicators such as click rates, conversion rates, and number of impressions.

[0784] "Means of collection" refers to the means for obtaining past advertising performance data from advertising distribution platforms and databases.

[0785] The "preprocessing means" refers to means for imputing missing values ​​in the collected data, filtering the data, and normalizing the data.

[0786] "Means for conducting AI analysis" means means for applying machine learning algorithms using pre-processed data to evaluate the impact of elements of advertising creative on performance metrics.

[0787] "Means for extracting points for improvement in advertising creative" refers to a means for identifying specific changes to improve the effectiveness of advertising based on the results of AI analysis.

[0788] The "means for generating a report" refers to a means for visually organizing the extracted improvement points and creating a report using charts and graphs.

[0789] The "means for providing to the user's terminal" refers to the means for providing the generated report in a form that can be viewed by the user, such as a PDF format or a web-based dashboard.

[0790] A "data analysis algorithm" is a mathematical model or computational method for finding patterns and trends in data and making predictions.

[0791] "Data reconciliation" is the process of converting data collected from different data sources into a unified format and making it consistent.

[0792] "Data standardization" is the process of scaling all numerical data to a certain range, typically normalizing it to the range 0 to 1.

[0793] This invention is a system aimed at efficiently improving advertising creatives. The system is mainly composed of components such as a server, a terminal, and a user.

[0794] Data collection and preprocessing

[0795] The server collects past advertising performance data from ad distribution platforms and databases. This data includes performance indicators such as click-through rates, conversion rates, and number of impressions. A specific example of this data collection is using the Google Ads API. The API is used to obtain advertising performance data for a specified period and store it in a database.

[0796] The server performs missing value imputation, filtering, and normalization on the collected data. This preprocessing is performed using the Python Pandas library. Specifically, the following operations are performed:

[0797] Use the mean or median to impute missing values.

[0798] To remove unnecessary data, filtering is performed according to specific conditions.

[0799] Scale all numeric data to the range 0 to 1.

[0800] AI analysis

[0801] The server runs machine learning algorithms on the preprocessed data, such as random forests and neural networks. Specifically, it uses the Scikit-learn library to train a random forest model to predict ad performance.

[0802] The server performs the following operations:

[0803] Using the Scikit-learn library, training data is input into the model and learning is performed.

[0804] Use the trained model to predict performance on new advertising data.

[0805] Extracting points for improvement

[0806] Based on the results of the AI ​​analysis, the server extracts specific points for improvement in the ad creative. From the analysis results, it identifies the factors that have the greatest impact on performance and forms specific improvement proposals based on those.

[0807] Specific suggestions include using bright colors and highly visible fonts. This information is provided as useful feedback to designers and marketers.

[0808] Generate and deliver reports

[0809] The server visually organizes the identified improvement points and generates a report using charts and graphs. This report is output in PDF format and a web-based dashboard format. The Matplotlib library is used to generate various graphs and convert them to PDF.

[0810] The server finally delivers this report to the user's device, either by emailing the generated PDF file or uploading it to a web dashboard where the user can log in and view it.

[0811] Specific examples

[0812] For example, let's say an e-commerce site runs a successful summer sales campaign with the following performance data:

[0813] Click-through rate: 5%

[0814] Conversion rate: 2%

[0815] Views: 100,000

[0816] In contrast, here's the data for a failed fall sales campaign ad:

[0817] Click-through rate: 1%

[0818] Conversion rate: 0.5%

[0819] Views: 80,000

[0820] The server collects this data and performs preprocessing. The AI ​​analysis reveals that bright colors and highly visible fonts were effective for summer advertising. The extracted improvement points are compiled into a report as specific suggestions, such as "brighten the colors of advertising images and change the text font to a larger, more visible one."

[0821] Prompt Sentence Examples

[0822] Users are prompted to take specific actions, such as, "Use bright colors and large, highly visible text fonts in your next ad campaign images to increase click-through rates and conversion rates."

[0823] This series of systems makes it possible to improve advertising return on investment (ROI).

[0824] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0825] Step 1: Data collection

[0826] The first thing the server does is collect past advertising performance data from ad distribution platforms and databases. Specifically, it uses the Google Ads API to obtain data such as click-through rates, conversion rates, and number of impressions for a specified period. The input data is the unprocessed raw data obtained from the API, and the output is past advertising performance data stored in the server's database.

[0827] Specific behavior:

[0828] Access the Google Ads API using an authentication token to retrieve the required data.

[0829] The acquired data is saved in a CSV file or database.

[0830] Step 2: Data Preprocessing

[0831] The server then preprocesses the collected data, which includes missing value imputation, data filtering, and normalization. The input to the preprocessing is the raw data collected, and the output is a consistent and accurate dataset that can be analyzed.

[0832] Specific behavior:

[0833] Create a data frame using Python's Pandas library.

[0834] Impute missing values ​​with the mean or median.

[0835] Filter data based on specific criteria to remove unwanted data.

[0836] Scales numeric data to the range 0 to 1.

[0837] Step 3: AI analysis

[0838] The server runs a machine learning algorithm on the preprocessed data. This step uses models like random forests or neural networks. The input is the preprocessed dataset, and the output is an analysis showing the impact of each element of the ad creative on performance metrics.

[0839] Specific behavior:

[0840] Train a random forest model using the Scikit-learn library.

[0841] Use the trained model to predict performance on new advertising data.

[0842] Step 4: Identifying areas for improvement

[0843] The server extracts specific points for improvement in the ad creative based on the results of the AI ​​analysis. The input is the AI ​​analysis results, and the output is specific improvement suggestions.

[0844] Specific behavior:

[0845] The factors with the greatest impact are extracted from the analysis results.

[0846] Create specific improvement suggestions for the use of colors in advertising images and text fonts.

[0847] Step 5: Generate reports

[0848] The server visually organizes the extracted improvement points and generates a report using charts and graphs. The input is the improvement proposals, and the output is a report in PDF format or a web dashboard format.

[0849] Specific behavior:

[0850] Generate various graphs using the Matplotlib library.

[0851] Convert the generated graphs to PDF and compile a report.

[0852] Step 6: Providing the report

[0853] The server provides the generated report to the user's device. This step involves emailing the PDF file or uploading it to a web dashboard. The input is the generated report and the output is the report provided to the user.

[0854] Specific behavior:

[0855] Generate a PDF report and send it to the user via email.

[0856] Upload reports to a web dashboard so users can log in and view them.

[0857] (Application example 1)

[0858] 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."

[0859] Conventional methods for improving advertising creatives require time and effort to collect and analyze past performance data, making it difficult to propose effective improvements in real time. Furthermore, since specific improvement proposals based on the expected effectiveness of advertising creatives cannot be obtained, there are limited ways to maximize advertising performance.

[0860] 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.

[0861] In this invention, the server includes means for collecting performance data of past advertisements, means for preprocessing the collected data, means for performing AI analysis based on the preprocessed data, means for extracting improvement points for the advertisement creative based on the results of the AI ​​analysis, means for generating a report including the improvement points, means for providing the generated report to a user's terminal, means for providing improvement suggestions for the advertisement creative in real time, and means for generating responsive prompt sentences. This enables efficient creative improvement in real time based on the performance data of advertisements, thereby maximizing advertisement performance.

[0862] "Past advertising performance data" refers to data including performance indicators related to past advertising distribution, such as click rates, conversion rates, and number of impressions of advertisements.

[0863] "Data preprocessing" is the process of completing the collected data by imputing missing values, filtering, and normalizing it, preparing it for analysis.

[0864] "AI analysis" is the process of analyzing data using machine learning algorithms to assess the impact that specific elements of ad creative have on performance metrics.

[0865] "Improvement points" are specific suggested changes to improve the performance of ad creatives, extracted based on the results of AI analysis.

[0866] "Report generation" is the process of visually organizing improvement points and outputting them in a format that is easy for users to understand using charts and graphs.

[0867] "User device" refers to an electronic device owned by a user, such as a smartphone, tablet, or PC, to which a report is provided.

[0868] "Means for providing real-time improvement suggestions for advertising creatives" is a function for quickly providing users with improvement suggestions for advertising creatives that are generated based on the results of AI analysis.

[0869] The "means for generating prompt sentences" is a function that automatically creates sentences to make suggestions or ask questions to the user in order to support the improvement of advertising creatives.

[0870] This invention is a system for realizing efficient improvement of advertising creatives, and is composed of the following components: The system includes the following elements: a server, a terminal, and a user.

[0871] Data collection and preprocessing

[0872] In the initial stage, the server collects historical ad performance data from ad serving platforms and databases. This data includes performance indicators such as click-through rates, conversion rates, and number of impressions. Once collected, the server fills in missing values, filters, and normalizes the data to prepare it for analysis. This ensures consistency and accuracy of the data.

[0873] AI analysis

[0874] After data preprocessing is complete, the server performs analysis using machine learning algorithms. Learning models used include random forests and neural networks. This allows the impact of specific elements of ad creative on performance indicators to be evaluated. AI analysis learns the characteristics of past successful and unsuccessful ads, allowing for prediction of the performance of new ads.

[0875] Extracting points for improvement

[0876] Based on the results of the AI ​​analysis, the server extracts specific points for improvement in the ad creative. For example, if it concludes that changes in color, text font, or layout will contribute to an increase in click-through rate, it will present that information as a specific improvement suggestion.

[0877] Generate and deliver reports

[0878] The server visually organizes the extracted improvement points and generates a report using charts and graphs. The report is output in PDF or web-based dashboard format and is designed to be easy for users to view. Finally, the server sends the generated report to the terminal, where the user can review it and take improvement actions.

[0879] Real-time improvement suggestions for advertising creatives

[0880] Based on AI analysis, the server generates real-time suggestions for improving ad creatives and provides them to users' devices. When users create new ads, they can make specific changes while receiving immediate feedback from the server.

[0881] Generate prompt statement

[0882] Furthermore, the server can generate prompt sentences to assist the user in creating an advertisement. The prompt sentences are sentences that suggest improvements to the advertisement creative and serve to prompt the user to take specific action to improve the advertisement.

[0883] Specific examples

[0884] For example, an e-commerce site may have a very successful summer sales campaign with the following performance data:

[0885] Click-through rate: 5%

[0886] Conversion rate: 2%

[0887] Views: 100,000

[0888] In contrast, here's the data for a failed fall sales campaign ad:

[0889] Click-through rate: 1%

[0890] Conversion rate: 0.5%

[0891] Views: 80,000

[0892] The server collects this data and performs preprocessing. The AI ​​analysis reveals that bright colors and highly visible fonts were effective for summer advertising. The extracted improvement points are compiled into a report as specific suggestions, such as "brighten the colors of advertising images and change the text font to a larger, more visible one." Finally, the server provides this report to the user's device, and the user can design the next advertising campaign based on the suggested improvements, which is expected to improve advertising efficiency.

[0893] Also, when a user creates a new ad creative, the server provides a prompt like this:

[0894] "We recommend using brighter colors and more legible fonts in your new ads. This change may increase your click-through rate."

[0895] This allows the system to help improve effective advertising creative in real time.

[0896] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0897] Step 1:

[0898] The server collects performance data of past ads from ad distribution platforms and databases. The collected data includes click-through rates, conversion rates, number of impressions, etc. The input is the performance data of the ads, and the output is the collected raw data.

[0899] Step 2:

[0900] The server preprocesses the collected data. Specifically, it imputes missing values ​​in the data, filters to remove noise, and scales and normalizes the data. The input is the raw data collected in step 1, and the output is the preprocessed data.

[0901] Step 3:

[0902] The server uses the preprocessed data to perform AI analysis, using machine learning models such as random forests and neural networks to evaluate the impact of specific elements of the ad creative on performance metrics. The input is the preprocessed data, and the output is the analysis results.

[0903] Step 4:

[0904] The server extracts specific points for improvement in the ad creative based on the results of the AI ​​analysis. For example, it generates specific suggestions, such as how changing the color usage or text font will contribute to improving click-through rates. The input is the analysis results obtained in step 3, and the output is the points for improvement.

[0905] Step 5:

[0906] The server generates a visually organized report using the extracted improvement points. This report is output in a user-friendly format using charts and graphs. The input is the improvement points extracted in step 4, and the output is the report.

[0907] Step 6:

[0908] The server provides the generated report to the user's device, allowing the user to review the report and take specific improvement actions. The input is the report generated in step 5, and the output is the report delivered to the user's device.

[0909] Step 7:

[0910] The server provides real-time suggestions for improving ad creatives. Users can receive immediate feedback when creating new ad creatives. The input is the initial data for new ad creatives and analysis results based on past performance data, and the output is real-time suggestions for improvement.

[0911] Step 8:

[0912] The server generates a prompt to assist the user in creating an ad. The prompt serves to encourage the user to take specific action to improve the ad. The input is the improvement proposal obtained in step 7, and the output is the prompt. A specific example would be, "We recommend that you use brighter colors in your new ad and change the font to one that is more visible. This change may improve your click-through rate."

[0913] 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.

[0914] This invention is a system aimed at efficiently improving advertising creatives, utilizing both past advertising performance data and user emotional data. The system is primarily comprised of a server, a terminal, and a user. Here, we will explain in detail how this system is implemented.

[0915] Data collection and preprocessing

[0916] At the initial stage, the server sends an API request to an ad serving platform or database to collect performance data of past ads. The collected data includes performance indicators such as click-through rate, conversion rate, and number of impressions. In addition, the server uses an emotion engine to obtain emotional data from the user's facial expressions, voice tone, text input, etc. This also collects the emotions of the user viewing the ad.

[0917] Data Preprocessing

[0918] The server preprocesses the advertising performance data and sentiment data collected. Specifically, it fills in missing values ​​in the data and performs filtering to maintain consistency. It also normalizes the data scale and prepares it for analysis. This makes both data suitable for analysis.

[0919] AI analysis

[0920] The server uses machine learning algorithms to analyze the preprocessed advertising performance data and emotion data. The learning models include random forests and neural networks, and evaluate the impact of each element of the advertising creative on performance indicators and user emotions. This analysis learns the characteristics of past successful and unsuccessful ads, making it possible to predict the performance and evaluate emotions of new ads.

[0921] Extracting points for improvement

[0922] Based on the results of the AI ​​analysis, the server extracts specific areas for improvement in the ad creative. For example, it identifies common characteristics of ads with high click-through rates and conversion rates, as well as the corresponding positive emotional responses, and summarizes these as areas for improvement. This information is presented in the form of suggestions for color usage, fonts, layout, etc.

[0923] Generate and deliver reports

[0924] The server visually organizes the extracted improvement points and generates a report using charts and graphs. The report is created in PDF or web-based dashboard format, emphasizing ease of viewing and understanding. It also includes the results of sentiment data analysis, showing how advertising affects users' emotions. Finally, the server provides the generated report to the user's device.

[0925] Specific examples

[0926] For example, let's say an e-commerce site's summer sales campaign was extremely successful, resulting in the following performance and sentiment data:

[0927] Click-through rate: 5%

[0928] Conversion rate: 2%

[0929] Views: 100,000

[0930] Positive emotional responses (smiles, positive comments): 70%

[0931] In contrast, here's the data for a failed fall sales campaign ad:

[0932] Click-through rate: 1%

[0933] Conversion rate: 0.5%

[0934] Views: 80,000

[0935] Negative emotional responses (sad facial expressions, negative comments): 50%

[0936] The server collects this data and performs preprocessing. The AI ​​analysis results reveal that bright colors and highly visible fonts are effective in summer advertisements and also evoke positive emotional responses from users. The extracted improvement points are compiled into a report as specific suggestions, such as "brighten the colors of the advertisement images and change the text font to a larger, more visible one."

[0937] Finally, the server provides this report to the user's device, and the next advertising campaign is designed based on the user's suggested improvements, which is expected to improve advertising efficiency. Through this series of processes, advertising return on investment (ROI) and advertising that takes user emotions into consideration are realized.

[0938] The processing flow will be explained below.

[0939] Step 1:

[0940] The server sends an API request to an ad serving platform or database to collect performance data of past ads. The collected data includes performance indicators such as click-through rate, conversion rate, and number of impressions. At the same time, an emotion engine is run to obtain emotional data from the user's facial expressions, tone of voice, text input, etc.

[0941] Step 2:

[0942] The server preprocesses the advertising performance data and sentiment data collected. First, if any missing values ​​exist, they are imputed appropriately. Next, the data is filtered to ensure consistency. Finally, the scale of the input data is normalized and converted into a format suitable for machine learning algorithms.

[0943] Step 3:

[0944] The server then feeds the pre-processed data into machine learning algorithms, such as random forests and neural networks, to evaluate the impact of each element of the ad creative on performance metrics and user sentiment.

[0945] Step 4:

[0946] The server trains the machine learning model and performs cross-validation, checking the accuracy of the model based on the results of the cross-validation and optimizing the model as needed, thereby building a more accurate predictive model.

[0947] Step 5:

[0948] Based on the analysis results, the server extracts specific points for improvement in the ad creative. For example, it identifies common characteristics of ads with high click-through rates and conversion rates, as well as elements that evoke positive emotional responses. This information is presented as improvement suggestions in the form of specific things like color usage, fonts, and layout.

[0949] Step 6:

[0950] The server visually organizes the identified improvement points and generates a report using charts and graphs in PDF or web-based dashboard format, which also includes the results of sentiment data analysis to make it easier to understand the overall picture.

[0951] Step 7:

[0952] The server sends the generated report to the user's device, allowing the user to review the report and use it to design and improve new ad creatives.

[0953] Step 8:

[0954] We carefully review the reports received by users and design and produce new ad creatives based on the suggested improvements. Based on the suggestions, we make specific improvements such as changes to color usage, layout, and text fonts.

[0955] Step 9:

[0956] The user then redistributes the improved ad, collecting new performance data. This new data is then sent back to the server for analysis in the next cycle. This process is repeated, resulting in continuous improvements in ad performance and user emotional response.

[0957] Example 2

[0958] 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."

[0959] Conventional advertising improvement systems have been unable to effectively improve advertising performance because they have been unable to fully integrate past advertising performance data and user emotion data, and their specific proposals for improving advertising creatives are abstract. Furthermore, it has been difficult to efficiently collect and analyze user emotion data and perform emotion evaluation based on advertising performance.

[0960] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting past advertising performance data, means for pre-processing the collected data, means for collecting user emotion data, means for performing AI analysis based on the pre-processed data, means for extracting improvement points for advertising creative based on the results of the AI ​​analysis, means for generating a report including the improvement points, and means for providing the generated report to the user's terminal. This makes it possible to comprehensively analyze past advertising performance data and user emotion data and effectively improve advertising performance through specific improvement suggestions.

[0961] "Advertising performance data" refers to performance indicators such as click rates, conversion rates, and number of impressions that indicate the effectiveness of past advertising campaigns.

[0962] "Preprocessing" refers to the process of filling in missing values ​​in the data, filtering to maintain consistency, and normalizing the data to prepare it for analysis.

[0963] "User emotion data" refers to data that indicates the user's emotional state, obtained from the user's facial expression, voice tone, text input, and the like.

[0964] "AI analytics" is the process of using machine learning algorithms to analyze pre-processed data and evaluate the impact of each element of advertising creative on performance metrics and user sentiment.

[0965] "Points for improving advertising creative" refers to specific changes and improvements that are needed to improve advertising performance based on the results of AI analysis.

[0966] A "report" is a report that visually organizes the analysis results and extracted improvement points and displays them using charts, graphs, etc.

[0967] A "user's terminal" is a device such as a computer, smartphone, or tablet that a user uses to receive reports provided by the server.

[0968] This invention is a system aimed at efficiently improving advertising creatives, and provides a means for integrating and analyzing past advertising performance data and user emotion data. The system is primarily comprised of components: a server, a terminal, and a user. Here, we will explain in detail how this system is implemented.

[0969] Data collection

[0970] The server first sends an API request to an ad serving platform or database to collect past ad performance data. Specific APIs typically used are the Google Ads API and Facebook Ads API. This data includes click-through rates, conversion rates, and number of impressions.

[0971] The server also collects the user's emotional data using an emotion engine, such as Microsoft Azure Emotion API or IBM Watson Tone Analyzer, which analyzes and obtains the user's emotional state from their facial expressions, voice tone, and text input. Emotional data is collected when the device captures the user's video and audio and sends them to the server.

[0972] Data Preprocessing

[0973] The server preprocesses the advertising performance data and sentiment data collected by the server. The Python pandas library is used to impute missing values, and data filtering and normalization are also performed. For example, this includes outlier detection and normalization of the data scale using a scaling library.

[0974] AI analysis

[0975] The server then analyzes the pre-processed data using machine learning algorithms, specifically models such as random forests and neural networks. The analysis uses the scikit-learn library to train and evaluate the models. The AI ​​analysis evaluates the impact of each element of the ad creative on performance metrics and user sentiment.

[0976] Extracting points for improvement

[0977] The server extracts specific areas for improvement in the ad creative based on the results of the AI ​​analysis. It identifies common characteristics of ads with high click-through rates and conversion rates and summarizes the areas for improvement. For example, it generates suggestions for "bright colors" and "highly visible fonts."

[0978] Generate and deliver reports

[0979] The server visually organizes the extracted improvement points and generates a report using charts and graphs. Matplotlib and Plotly are used to generate the report, which is automatically created using a PDF generator. Finally, the server provides the generated report to the user's device. The report is delivered as a download link or email attachment, and the user can use it to design their next advertising campaign.

[0980] Specific examples

[0981] For example, suppose an e-commerce site runs a summer sale campaign ad with a 5% click-through rate, a 2% conversion rate, 100,000 impressions, and a 70% positive emotional response. In contrast, if a fall sale campaign ad has a 1% click-through rate, a 0.5% conversion rate, 80,000 impressions, and a 50% negative emotional response, the server collects and preprocesses this data and performs AI analysis. The analysis results suggest that bright colors and highly visible fonts are effective for summer ads.

[0982] The extracted improvement points are compiled into a report with specific suggestions, such as "Use brighter colors in the ad images and change the text font to a larger, more visible one." Finally, the server provides the generated report to the user's device, allowing the user to incorporate the improvements into their next ad campaign.

[0983] Prompt Sentence Examples

[0984] Please suggest areas for improvement in the ad creative based on the advertising data below.

[0985] (Successful summer sale campaign advertising data)

[0986] Click-through rate: 5%

[0987] Conversion rate: 2%

[0988] Views: 100,000

[0989] Positive emotional response: 70%

[0990] (Failed Fall Sale Campaign Advertising Data)

[0991] Click-through rate: 1%

[0992] Conversion rate: 0.5%

[0993] Views: 80,000

[0994] Negative emotional reactions: 50%

[0995] suggestion:

[0996] Use bright colors and highly visible fonts, just like in summer ads, to increase click-through rates and conversion rates.

[0997] With the above configuration, this system improves advertising performance and proposes advertising creatives that take user emotions into consideration.

[0998] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0999] Step 1: Collect advertising performance data

[1000] The server sends API requests to ad distribution platforms and databases to collect past ad performance data. Specifically, it uses the Google Ads API and Facebook Ads API. The input is API authentication information and the query required to obtain data, and the output is data on ad performance indicators such as click-through rate (CTR), conversion rate (CVR), and number of impressions. The server performs API authentication using the client ID and secret, and obtains data from the specified endpoint.

[1001] Step 2: Collecting emotion data

[1002] The server runs an emotion engine and collects emotion data from the user's facial expressions, voice tone, and text input. The inputs are camera footage, audio clips, and text data, and the emotional state is analyzed based on this. Specifically, the Microsoft Azure Emotion API and IBM Watson Tone Analyzer are used. The device captures the user's video and audio and sends them to the server to obtain emotion data. The output is data indicating the user's emotional state (e.g., smile rate, negative comment rate).

[1003] Step 3: Preprocessing the data

[1004] The advertising performance data and sentiment data collected by the server are preprocessed. The input is the raw data obtained in steps 1 and 2. Specific preprocessing includes missing value imputation, filtering, and normalization. The Python pandas library is used to impute missing values ​​with the mean or median, and outliers are filtered. Furthermore, a scaling library is used to normalize the data. The output is data that is ready for analysis.

[1005] Step 4: AI analysis

[1006] The server applies a machine learning algorithm to the preprocessed data and performs analysis. The input is the preprocessed advertising performance data and emotion data. Specifically, a random forest or neural network model is used, and the model is trained and evaluated using the scikit-learn library. The output is an evaluation of the impact of each element of the advertising creative on performance indicators and user emotions. The server randomly shuffles the training data and divides it into a training set and a test set. The algorithm is applied to train the model, and accuracy is verified on the test set.

[1007] Step 5: Identifying areas for improvement

[1008] The server extracts specific improvement points for the ad creative based on the results of the AI ​​analysis. The input is the data results of the AI ​​analysis. The server analyzes the importance of the features and displays a ranking of the elements that have the greatest impact on performance. The improvement points are written out as specific suggestions using natural language generation (NLG) technology. The output is specific changes and improvements (e.g., bright colors, highly visible fonts) as improvement suggestions.

[1009] Step 6: Generate reports

[1010] The server visually organizes the improvement points extracted and generates a report using charts and graphs. The input is improvement point data. Matplotlib and Plotly are used to generate the report. A data visualization library is used to graphically represent each indicator, and a PDF generator is used to automatically generate the report. The output is a report in PDF format or a web-based dashboard format.

[1011] Step 7: Providing the report

[1012] The server provides the generated report to the user's device. The input is the generated report data. The server configures the SMTP settings to send the report to the user's email address, and includes links and attachments in the email body. The output is a report provided to the user's device, which the user can use to design their next advertising campaign.

[1013] (Application example 2)

[1014] 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."

[1015] Conventional systems for improving advertising creatives rely solely on past performance data and are unable to consider the emotional responses of viewers. This makes it difficult to grasp the specific emotional impact that advertising has on viewers and to propose appropriate improvements based on that. In particular, there was a need for a method to evaluate the impact of visual and audio factors on viewer emotions, in addition to click-through rates and conversion rates.

[1016] 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.

[1017] In this invention, the server includes means for collecting performance data of past advertisements, means for collecting emotional response data, means for preprocessing the collected performance data and emotional response data, means for performing AI analysis based on the preprocessed data and emotional response data, means for extracting improvement points for the advertisement creative based on the results of the AI ​​analysis, means for generating a report including the improvement points, and means for providing the generated report to the user's connection device, thereby enabling efficient improvement of the advertisement creative based on the viewer's emotional response.

[1018] "Past advertising performance data" refers to data including performance indicators such as click rates, conversion rates, and number of impressions obtained in ad distribution.

[1019] "Preprocessing" refers to the process of preparing data suitable for analysis after collection by performing tasks such as filling in missing values, filtering, and normalizing the scale.

[1020] "Emotional response data" refers to emotional data of a user analyzed from facial expressions, tone of voice, text input, and the like while the user is viewing an advertisement.

[1021] "AI analysis" is the process of using machine learning algorithms to evaluate the impact of each element of advertising creative on performance metrics and emotions based on past advertising performance data and emotional response data.

[1022] "Ad creative" refers to the design elements of an ad, such as images, text, layout, and color usage.

[1023] "Improvement points" are specific elements or methods that are recommended to be changed or improved based on the results of AI analysis in order to improve the performance of advertising creatives.

[1024] "Report generation" is the process of visually organizing the AI ​​analysis results and points for improvement, and creating a report using charts and graphs to provide to the user.

[1025] "User's connected device" refers to the device (e.g., smartphone, tablet, PC) used by the user to provide the above-mentioned report.

[1026] A "machine learning algorithm" is an algorithm (e.g., random forest, neural network) that uses large amounts of data to train a model and make predictions or classifications from new data.

[1027] This invention is a system for efficiently improving advertising creatives, which utilizes a combination of past advertising performance data and user emotional response data. Specifically, it includes the following components:

[1028] First, the server uses an API to collect past advertising performance data from the ad distribution platform. This data includes performance indicators such as click-through rate, conversion rate, and number of impressions. Furthermore, a sentiment analysis engine is used to collect emotional responses from users' facial expressions, voice tone, text input, etc. This allows the server to also collect emotional data on users who view ads.

[1029] Next, the server preprocesses the collected ad performance data and emotional response data. Specifically, it imputes missing values, filters to maintain consistency, and normalizes the data scale. This prepares both data sets for analysis. Data preprocessing is performed using data processing libraries such as Pandas and Scikit-learn.

[1030] The server then uses machine learning algorithms to analyze the pre-processed ad performance data and emotional response data. At this stage, it uses tools such as random forests and neural networks (using TensorFlow and Keras) to create models that evaluate the impact of each element of the ad creative on performance indicators and emotions.

[1031] Based on the analysis results, the server can identify areas for improvement in ad creatives. For example, it may determine that ads with high click-through rates and conversion rates are effective with specific colors, fonts, and layouts, and that they also evoke positive emotional responses from users.

[1032] Points for improvement are compiled into specific proposals, which are then generated by the server as a report. The report is visually organized using charts and graphs and delivered to the user's connected device (smartphone, tablet, or PC). Visualization libraries such as Matplotlib and Seaborn are used to generate the report. Reports can also be created in PDF format using ReportLab.

[1033] As a specific example, consider a case where a server collects and analyzes the following data:

[1034] Summer sale advertisement:

[1035] Click-through rate: 5%

[1036] Conversion rate: 2%

[1037] Views: 100,000

[1038] Positive emotional responses (smiles, positive comments): 70%

[1039] Fall sale ad:

[1040] Click-through rate: 1%

[1041] Conversion rate: 0.5%

[1042] Views: 80,000

[1043] Negative emotional responses (sad facial expressions, negative comments): 50%

[1044] An example of a prompt sentence is as follows:

[1045] System Input:

[1046] The system collects data for a summer sale ad with a 5% click-through rate, a 2% conversion rate, 100,000 impressions, and a 70% positive emotional response, and a fall sale ad with a 1% click-through rate, a 0.5% conversion rate, 80,000 impressions, and a 50% negative emotional response.

[1047] Inference result:

[1048] We suggest that brightening the colors of advertising images and enlarging the font size would be effective.

[1049] This will enable efficient improvement of advertising creatives based on viewers' emotional responses.

[1050] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1051] Step 1:

[1052] The server uses the API from the ad distribution platform to collect data on past ad performance and user emotional response. This data includes click-through rates, conversion rates, number of impressions, users' facial expressions, voice tone, and text input. This data is the basis for measuring the effectiveness of ads and determining users' emotional responses.

[1053] Input: API of ad distribution platform, user facial expressions, voice tone, text input

[1054] Output: Collected advertising performance data and emotional response data

[1055] Step 2:

[1056] The ad performance data and emotional response data collected by the server are preprocessed. Specifically, missing values ​​are imputed, inconsistent data is filtered, and the data scale is normalized. Data processing libraries such as Pandas and Scikit-learn are used for preprocessing. Preprocessing improves data consistency and suitability for analysis.

[1057] Input: Collected advertising performance data and emotional response data

[1058] Output: Preprocessed ad performance data and emotional response data

[1059] Step 3:

[1060] The server performs AI analysis based on the preprocessed ad performance data and emotional response data. This analysis applies machine learning algorithms such as random forests and neural networks (using TensorFlow, Keras, etc.). The AI ​​analysis evaluates the impact of ad creative elements on performance indicators and user emotional responses and creates a model.

[1061] Input: Preprocessed advertising performance data and emotional response data

[1062] Output: A model for evaluating the impact of advertising creative elements on performance metrics and emotional responses.

[1063] Step 4:

[1064] Based on the results of the AI ​​analysis, the server extracts points for improvement in the ad creative. For example, it suggests specific ways to improve elements such as color usage, font, size, and placement. At this stage, feedback is obtained to suggest effective improvement measures based on the analysis results.

[1065] Input: A model for evaluating the impact of advertising creative elements on performance indicators and emotional responses.

[1066] Output: Specific suggestions for improvement

[1067] Step 5:

[1068] The server visually organizes the identified improvement points and generates a report, which can be created in PDF or web-based dashboard format and designed to include charts and graphs using visualization libraries such as Matplotlib, Seaborn, and ReportLab.

[1069] Input: Specific suggestions for improvement

[1070] Output: Visualized report

[1071] Step 6:

[1072] The server provides the generated report to the user's connected device (smartphone, tablet, PC), allowing the user to improve their ad creative and apply it to their next ad campaign to maximize the effectiveness of their ads.

[1073] Input: Visualized report

[1074] Output: Report provided to user's connected devices

[1075] 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.

[1076] 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.

[1077] 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.

[1078] [Fourth embodiment]

[1079] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1080] 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.

[1081] 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).

[1082] 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.

[1083] 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.

[1084] 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).

[1085] 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. 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.

[1086] 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.

[1087] 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.

[1088] 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.

[1089] 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.

[1090] 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.

[1091] 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."

[1092] The present invention is a system for efficiently improving advertising creatives, and is specifically implemented as follows: The system is mainly composed of components of a server, a terminal, and a user.

[1093] Data collection and preprocessing

[1094] In the initial stage, the server collects historical ad performance data from ad serving platforms and databases. This data includes performance indicators such as click-through rates, conversion rates, and number of impressions. Once collected, the server fills in missing values, filters, and normalizes the data to prepare it for analysis. This ensures consistency and accuracy of the data.

[1095] AI analysis

[1096] The server then uses machine learning algorithms to analyze the pre-processed data. Learning models include random forests and neural networks, and evaluate the impact that specific elements of ad creative have on performance indicators. This analysis allows the system to learn the characteristics of past successful and unsuccessful ads, making it possible to predict the performance of new ads.

[1097] Extracting points for improvement

[1098] Based on the results of AI analysis, the server extracts specific points for improvement in the ad creative. For example, if it concludes that changes in color, text font, or layout will contribute to an increase in click-through rate, it presents that information as a specific improvement suggestion, providing designers and marketers with clear guidelines.

[1099] Generate and deliver reports

[1100] The server visually organizes the identified improvement points and generates a report using charts and graphs. The report is output in PDF or web-based dashboard format and is designed for easy viewing by the user. Finally, the server sends the generated report to the terminal, where the user can review it and take improvement actions.

[1101] Specific examples

[1102] For example, let's say an e-commerce site has a highly successful summer sales campaign with the following performance data:

[1103] Click-through rate: 5%

[1104] Conversion rate: 2%

[1105] Views: 100,000

[1106] In contrast, here's the data for a failed fall sales campaign ad:

[1107] Click-through rate: 1%

[1108] Conversion rate: 0.5%

[1109] Views: 80,000

[1110] The server collects this data and performs preprocessing. The AI ​​analysis reveals that bright colors and highly visible fonts were effective for summer advertising. The extracted improvement points are compiled into a report as specific suggestions, such as "brighten the colors of advertising images and change the text font to a larger, more visible one."

[1111] Finally, the server provides this report to the user's device, and the user can design the next advertising campaign based on the suggested improvements, which can improve advertising efficiency. Through this process, an improvement in advertising return on investment (ROI) can be realized.

[1112] The processing flow will be explained below.

[1113] Step 1:

[1114] The server sends an API request to an ad serving platform or database to collect performance data for past ads, including performance metrics such as click-through rates, conversion rates, and impressions.

[1115] Step 2:

[1116] The server preprocesses the collected data by imputing missing values, filtering for consistency, and normalizing the data scale, making the data suitable for analysis.

[1117] Step 3:

[1118] The server then feeds the pre-processed data into machine learning algorithms, such as models like random forests or neural networks, that analyze the impact of each element of the ad creative on performance.

[1119] Step 4:

[1120] The server trains the machine learning model and checks the model's predictive accuracy through cross-validation, thereby building an accurate predictive model.

[1121] Step 5:

[1122] The server extracts specific points for improvement based on the analysis results. For example, it identifies common characteristics of ads with high click-through rates and conversion rates and summarizes them as areas for improvement. This information is presented in the form of suggestions for color, font, layout, etc.

[1123] Step 6:

[1124] The server visually organizes the identified improvement points and generates a report using charts and graphs. The report is created in PDF or web-based dashboard format, with an emphasis on ease of viewing and understanding.

[1125] Step 7:

[1126] The server sends the generated report to the user's terminal, which allows the user to quickly take concrete action regarding the creation of new advertisements.

[1127] Step 8:

[1128] We review the reports received by the user and design and produce new ad creatives based on the suggested improvements. Based on the suggestions, we make specific improvements such as changes to color usage, layout, and text fonts.

[1129] Step 9:

[1130] The user then redistributes the improved ad, collecting new performance data. This data is then sent back to the server for analysis in the next cycle. By repeating this process, ad performance continues to improve.

[1131] Example 1

[1132] 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."

[1133] Improving the performance of advertising creatives is important in marketing, but conventional methods have made it difficult to quickly and accurately identify effective areas for improvement. To solve this problem, there is a need to develop a system that can efficiently and accurately extract and suggest areas for improvement using past advertising performance data.

[1134] 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.

[1135] In this invention, the server includes a means for collecting performance data of past advertisements, a means for preprocessing the collected data, and a means for performing AI analysis based on the preprocessed data, which makes it possible to evaluate the impact of elements of advertising creatives on performance indicators and quickly and accurately extract areas for improvement.

[1136] "Past advertising performance data" refers to data that includes advertising performance indicators such as click rates, conversion rates, and number of impressions.

[1137] "Means of collection" refers to the means for obtaining past advertising performance data from advertising distribution platforms and databases.

[1138] The "preprocessing means" refers to means for imputing missing values ​​in the collected data, filtering the data, and normalizing the data.

[1139] "Means for conducting AI analysis" means means for applying machine learning algorithms using pre-processed data to evaluate the impact of elements of advertising creative on performance metrics.

[1140] "Means for extracting points for improvement in advertising creative" refers to a means for identifying specific changes to improve the effectiveness of advertising based on the results of AI analysis.

[1141] The "means for generating a report" refers to a means for visually organizing the extracted improvement points and creating a report using charts and graphs.

[1142] The "means for providing to the user's terminal" refers to the means for providing the generated report in a form that can be viewed by the user, such as a PDF format or a web-based dashboard.

[1143] A "data analysis algorithm" is a mathematical model or computational method for finding patterns and trends in data and making predictions.

[1144] "Data reconciliation" is the process of converting data collected from different data sources into a unified format and making it consistent.

[1145] "Data standardization" is the process of scaling all numerical data to a certain range, typically normalizing it to the range 0 to 1.

[1146] This invention is a system aimed at efficiently improving advertising creatives. The system is mainly composed of components such as a server, a terminal, and a user.

[1147] Data collection and preprocessing

[1148] The server collects past advertising performance data from ad distribution platforms and databases. This data includes performance indicators such as click-through rates, conversion rates, and number of impressions. A specific example of this data collection is using the Google Ads API. The API is used to obtain advertising performance data for a specified period and store it in a database.

[1149] The server performs missing value imputation, filtering, and normalization on the collected data. This preprocessing is performed using the Python Pandas library. Specifically, the following operations are performed:

[1150] Use the mean or median to impute missing values.

[1151] To remove unnecessary data, filtering is performed according to specific conditions.

[1152] Scale all numeric data to the range 0 to 1.

[1153] AI analysis

[1154] The server runs machine learning algorithms on the preprocessed data, such as random forests and neural networks. Specifically, it uses the Scikit-learn library to train a random forest model to predict ad performance.

[1155] The server performs the following operations:

[1156] Using the Scikit-learn library, training data is input into the model and learning is performed.

[1157] Use the trained model to predict performance on new advertising data.

[1158] Extracting points for improvement

[1159] Based on the results of the AI ​​analysis, the server extracts specific points for improvement in the ad creative. From the analysis results, it identifies the factors that have the greatest impact on performance and forms specific improvement proposals based on those.

[1160] Specific suggestions include using bright colors and highly visible fonts. This information is provided as useful feedback to designers and marketers.

[1161] Generate and deliver reports

[1162] The server visually organizes the identified improvement points and generates a report using charts and graphs. This report is output in PDF format and a web-based dashboard format. The Matplotlib library is used to generate various graphs and convert them to PDF.

[1163] The server finally delivers this report to the user's device, either by emailing the generated PDF file or uploading it to a web dashboard where the user can log in and view it.

[1164] Specific examples

[1165] For example, let's say an e-commerce site runs a successful summer sales campaign with the following performance data:

[1166] Click-through rate: 5%

[1167] Conversion rate: 2%

[1168] Views: 100,000

[1169] In contrast, here's the data for a failed fall sales campaign ad:

[1170] Click-through rate: 1%

[1171] Conversion rate: 0.5%

[1172] Views: 80,000

[1173] The server collects this data and performs preprocessing. The AI ​​analysis reveals that bright colors and highly visible fonts were effective for summer advertising. The extracted improvement points are compiled into a report as specific suggestions, such as "brighten the colors of advertising images and change the text font to a larger, more visible one."

[1174] Prompt Sentence Examples

[1175] Users are prompted to take specific actions, such as, "Use bright colors and large, highly visible text fonts in your next ad campaign images to increase click-through rates and conversion rates."

[1176] This series of systems makes it possible to improve advertising return on investment (ROI).

[1177] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1178] Step 1: Data collection

[1179] The first thing the server does is collect past advertising performance data from ad distribution platforms and databases. Specifically, it uses the Google Ads API to obtain data such as click-through rates, conversion rates, and number of impressions for a specified period. The input data is the unprocessed raw data obtained from the API, and the output is past advertising performance data stored in the server's database.

[1180] Specific behavior:

[1181] Access the Google Ads API using an authentication token to retrieve the required data.

[1182] The acquired data is saved in a CSV file or database.

[1183] Step 2: Data Preprocessing

[1184] The server then preprocesses the collected data, which includes missing value imputation, data filtering, and normalization. The input to the preprocessing is the raw data collected, and the output is a consistent and accurate dataset that can be analyzed.

[1185] Specific behavior:

[1186] Create a data frame using Python's Pandas library.

[1187] Impute missing values ​​with the mean or median.

[1188] Filter data based on specific criteria to remove unwanted data.

[1189] Scales numeric data to the range 0 to 1.

[1190] Step 3: AI analysis

[1191] The server runs a machine learning algorithm on the preprocessed data. This step uses models like random forests or neural networks. The input is the preprocessed dataset, and the output is an analysis showing the impact of each element of the ad creative on performance metrics.

[1192] Specific behavior:

[1193] Train a random forest model using the Scikit-learn library.

[1194] Use the trained model to predict performance on new advertising data.

[1195] Step 4: Identifying areas for improvement

[1196] The server extracts specific points for improvement in the ad creative based on the results of the AI ​​analysis. The input is the AI ​​analysis results, and the output is specific improvement suggestions.

[1197] Specific behavior:

[1198] The factors with the greatest impact are extracted from the analysis results.

[1199] Create specific improvement suggestions for the use of colors in advertising images and text fonts.

[1200] Step 5: Generate reports

[1201] The server visually organizes the extracted improvement points and generates a report using charts and graphs. The input is the improvement proposals, and the output is a report in PDF format or a web dashboard format.

[1202] Specific behavior:

[1203] Generate various graphs using the Matplotlib library.

[1204] Convert the generated graphs to PDF and compile a report.

[1205] Step 6: Providing the report

[1206] The server provides the generated report to the user's device. This step involves emailing the PDF file or uploading it to a web dashboard. The input is the generated report and the output is the report provided to the user.

[1207] Specific behavior:

[1208] Generate a PDF report and send it to the user via email.

[1209] Upload reports to a web dashboard so users can log in and view them.

[1210] (Application example 1)

[1211] 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."

[1212] Conventional methods for improving advertising creatives require time and effort to collect and analyze past performance data, making it difficult to propose effective improvements in real time. Furthermore, since specific improvement proposals based on the expected effectiveness of advertising creatives cannot be obtained, there are limited ways to maximize advertising performance.

[1213] 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.

[1214] In this invention, the server includes means for collecting performance data of past advertisements, means for preprocessing the collected data, means for performing AI analysis based on the preprocessed data, means for extracting improvement points for the advertisement creative based on the results of the AI ​​analysis, means for generating a report including the improvement points, means for providing the generated report to a user's terminal, means for providing improvement suggestions for the advertisement creative in real time, and means for generating responsive prompt sentences. This enables efficient creative improvement in real time based on the performance data of advertisements, thereby maximizing advertisement performance.

[1215] "Past advertising performance data" refers to data including performance indicators related to past advertising distribution, such as click rates, conversion rates, and number of impressions of advertisements.

[1216] "Data preprocessing" is the process of completing the collected data by imputing missing values, filtering, and normalizing it, preparing it for analysis.

[1217] "AI analysis" is the process of analyzing data using machine learning algorithms to assess the impact that specific elements of ad creative have on performance metrics.

[1218] "Improvement points" are specific suggested changes to improve the performance of ad creatives, extracted based on the results of AI analysis.

[1219] "Report generation" is the process of visually organizing improvement points and outputting them in a format that is easy for users to understand using charts and graphs.

[1220] "User device" refers to an electronic device owned by a user, such as a smartphone, tablet, or PC, to which a report is provided.

[1221] "Means for providing real-time improvement suggestions for advertising creatives" is a function for quickly providing users with improvement suggestions for advertising creatives that are generated based on the results of AI analysis.

[1222] The "means for generating prompt sentences" is a function that automatically creates sentences to make suggestions or ask questions to the user in order to support the improvement of advertising creatives.

[1223] This invention is a system for realizing efficient improvement of advertising creatives, and is composed of the following components: The system includes the following elements: a server, a terminal, and a user.

[1224] Data collection and preprocessing

[1225] In the initial stage, the server collects historical ad performance data from ad serving platforms and databases. This data includes performance indicators such as click-through rates, conversion rates, and number of impressions. Once collected, the server fills in missing values, filters, and normalizes the data to prepare it for analysis. This ensures consistency and accuracy of the data.

[1226] AI analysis

[1227] After data preprocessing is complete, the server performs analysis using machine learning algorithms. Learning models used include random forests and neural networks. This allows the impact of specific elements of ad creative on performance indicators to be evaluated. AI analysis learns the characteristics of past successful and unsuccessful ads, allowing for prediction of the performance of new ads.

[1228] Extracting points for improvement

[1229] Based on the results of the AI ​​analysis, the server extracts specific points for improvement in the ad creative. For example, if it concludes that changes in color, text font, or layout will contribute to an increase in click-through rate, it will present that information as a specific improvement suggestion.

[1230] Generate and deliver reports

[1231] The server visually organizes the extracted improvement points and generates a report using charts and graphs. The report is output in PDF or web-based dashboard format and is designed to be easy for users to view. Finally, the server sends the generated report to the terminal, where the user can review it and take improvement actions.

[1232] Real-time improvement suggestions for advertising creatives

[1233] Based on AI analysis, the server generates real-time suggestions for improving ad creatives and provides them to users' devices. When users create new ads, they can make specific changes while receiving immediate feedback from the server.

[1234] Generate prompt statement

[1235] Furthermore, the server can generate prompt sentences to assist the user in creating an advertisement. The prompt sentences are sentences that suggest improvements to the advertisement creative and serve to prompt the user to take specific action to improve the advertisement.

[1236] Specific examples

[1237] For example, an e-commerce site may have a very successful summer sales campaign with the following performance data:

[1238] Click-through rate: 5%

[1239] Conversion rate: 2%

[1240] Views: 100,000

[1241] In contrast, here's the data for a failed fall sales campaign ad:

[1242] Click-through rate: 1%

[1243] Conversion rate: 0.5%

[1244] Views: 80,000

[1245] The server collects this data and performs preprocessing. The AI ​​analysis reveals that bright colors and highly visible fonts were effective for summer advertising. The extracted improvement points are compiled into a report as specific suggestions, such as "brighten the colors of advertising images and change the text font to a larger, more visible one." Finally, the server provides this report to the user's device, and the user can design the next advertising campaign based on the suggested improvements, which is expected to improve advertising efficiency.

[1246] Also, when a user creates a new ad creative, the server provides a prompt like this:

[1247] "We recommend using brighter colors and more legible fonts in your new ads. This change may increase your click-through rate."

[1248] This allows the system to help improve effective advertising creative in real time.

[1249] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1250] Step 1:

[1251] The server collects performance data of past ads from ad distribution platforms and databases. The collected data includes click-through rates, conversion rates, number of impressions, etc. The input is the performance data of the ads, and the output is the collected raw data.

[1252] Step 2:

[1253] The server preprocesses the collected data. Specifically, it imputes missing values ​​in the data, filters to remove noise, and scales and normalizes the data. The input is the raw data collected in step 1, and the output is the preprocessed data.

[1254] Step 3:

[1255] The server uses the preprocessed data to perform AI analysis, using machine learning models such as random forests and neural networks to evaluate the impact of specific elements of the ad creative on performance metrics. The input is the preprocessed data, and the output is the analysis results.

[1256] Step 4:

[1257] The server extracts specific points for improvement in the ad creative based on the results of the AI ​​analysis. For example, it generates specific suggestions, such as how changing the color usage or text font will contribute to improving click-through rates. The input is the analysis results obtained in step 3, and the output is the points for improvement.

[1258] Step 5:

[1259] The server generates a visually organized report using the extracted improvement points. This report is output in a user-friendly format using charts and graphs. The input is the improvement points extracted in step 4, and the output is the report.

[1260] Step 6:

[1261] The server provides the generated report to the user's device, allowing the user to review the report and take specific improvement actions. The input is the report generated in step 5, and the output is the report delivered to the user's device.

[1262] Step 7:

[1263] The server provides real-time suggestions for improving ad creatives. Users can receive immediate feedback when creating new ad creatives. The input is the initial data for new ad creatives and analysis results based on past performance data, and the output is real-time suggestions for improvement.

[1264] Step 8:

[1265] The server generates a prompt to assist the user in creating an ad. The prompt serves to encourage the user to take specific action to improve the ad. The input is the improvement proposal obtained in step 7, and the output is the prompt. A specific example would be, "We recommend that you use brighter colors in your new ad and change the font to one that is more visible. This change may improve your click-through rate."

[1266] 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.

[1267] This invention is a system aimed at efficiently improving advertising creatives, utilizing both past advertising performance data and user emotional data. The system is primarily comprised of a server, a terminal, and a user. Here, we will explain in detail how this system is implemented.

[1268] Data collection and preprocessing

[1269] At the initial stage, the server sends an API request to an ad serving platform or database to collect performance data of past ads. The collected data includes performance indicators such as click-through rate, conversion rate, and number of impressions. In addition, the server uses an emotion engine to obtain emotional data from the user's facial expressions, voice tone, text input, etc. This also collects the emotions of the user viewing the ad.

[1270] Data Preprocessing

[1271] The server preprocesses the advertising performance data and sentiment data collected. Specifically, it fills in missing values ​​in the data and performs filtering to maintain consistency. It also normalizes the data scale and prepares it for analysis. This makes both data suitable for analysis.

[1272] AI analysis

[1273] The server uses machine learning algorithms to analyze the preprocessed advertising performance data and emotion data. The learning models include random forests and neural networks, and evaluate the impact of each element of the advertising creative on performance indicators and user emotions. This analysis learns the characteristics of past successful and unsuccessful ads, making it possible to predict the performance and evaluate emotions of new ads.

[1274] Extracting points for improvement

[1275] Based on the results of the AI ​​analysis, the server extracts specific areas for improvement in the ad creative. For example, it identifies common characteristics of ads with high click-through rates and conversion rates, as well as the corresponding positive emotional responses, and summarizes these as areas for improvement. This information is presented in the form of suggestions for color usage, fonts, layout, etc.

[1276] Generate and deliver reports

[1277] The server visually organizes the extracted improvement points and generates a report using charts and graphs. The report is created in PDF or web-based dashboard format, emphasizing ease of viewing and understanding. It also includes the results of sentiment data analysis, showing how advertising affects users' emotions. Finally, the server provides the generated report to the user's device.

[1278] Specific examples

[1279] For example, let's say an e-commerce site's summer sales campaign was extremely successful, resulting in the following performance and sentiment data:

[1280] Click-through rate: 5%

[1281] Conversion rate: 2%

[1282] Views: 100,000

[1283] Positive emotional responses (smiles, positive comments): 70%

[1284] In contrast, here's the data for a failed fall sales campaign ad:

[1285] Click-through rate: 1%

[1286] Conversion rate: 0.5%

[1287] Views: 80,000

[1288] Negative emotional responses (sad facial expressions, negative comments): 50%

[1289] The server collects this data and performs preprocessing. The AI ​​analysis results reveal that bright colors and highly visible fonts are effective in summer advertisements and also evoke positive emotional responses from users. The extracted improvement points are compiled into a report as specific suggestions, such as "brighten the colors of the advertisement images and change the text font to a larger, more visible one."

[1290] Finally, the server provides this report to the user's device, and the next advertising campaign is designed based on the user's suggested improvements, which is expected to improve advertising efficiency. Through this series of processes, advertising return on investment (ROI) and advertising that takes user emotions into consideration are realized.

[1291] The processing flow will be explained below.

[1292] Step 1:

[1293] The server sends an API request to an ad serving platform or database to collect performance data of past ads. The collected data includes performance indicators such as click-through rate, conversion rate, and number of impressions. At the same time, an emotion engine is run to obtain emotional data from the user's facial expressions, tone of voice, text input, etc.

[1294] Step 2:

[1295] The server preprocesses the advertising performance data and sentiment data collected. First, if any missing values ​​exist, they are imputed appropriately. Next, the data is filtered to ensure consistency. Finally, the scale of the input data is normalized and converted into a format suitable for machine learning algorithms.

[1296] Step 3:

[1297] The server then feeds the pre-processed data into machine learning algorithms, such as random forests and neural networks, to evaluate the impact of each element of the ad creative on performance metrics and user sentiment.

[1298] Step 4:

[1299] The server trains the machine learning model and performs cross-validation, checking the accuracy of the model based on the results of the cross-validation and optimizing the model as needed, thereby building a more accurate predictive model.

[1300] Step 5:

[1301] Based on the analysis results, the server extracts specific points for improvement in the ad creative. For example, it identifies common characteristics of ads with high click-through rates and conversion rates, as well as elements that evoke positive emotional responses. This information is presented as improvement suggestions in the form of specific things like color usage, fonts, and layout.

[1302] Step 6:

[1303] The server visually organizes the identified improvement points and generates a report using charts and graphs in PDF or web-based dashboard format, which also includes the results of sentiment data analysis to make it easier to understand the overall picture.

[1304] Step 7:

[1305] The server sends the generated report to the user's device, allowing the user to review the report and use it to design and improve new ad creatives.

[1306] Step 8:

[1307] We carefully review the reports received by users and design and produce new ad creatives based on the suggested improvements. Based on the suggestions, we make specific improvements such as changes to color usage, layout, and text fonts.

[1308] Step 9:

[1309] The user then redistributes the improved ad, collecting new performance data. This new data is then sent back to the server for analysis in the next cycle. This process is repeated, resulting in continuous improvements in ad performance and user emotional response.

[1310] Example 2

[1311] 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."

[1312] Conventional advertising improvement systems have been unable to effectively improve advertising performance because they have been unable to fully integrate past advertising performance data and user emotion data, and their specific proposals for improving advertising creatives are abstract. Furthermore, it has been difficult to efficiently collect and analyze user emotion data and perform emotion evaluation based on advertising performance.

[1313] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting past advertising performance data, means for pre-processing the collected data, means for collecting user emotion data, means for performing AI analysis based on the pre-processed data, means for extracting improvement points for advertising creative based on the results of the AI ​​analysis, means for generating a report including the improvement points, and means for providing the generated report to the user's terminal. This makes it possible to comprehensively analyze past advertising performance data and user emotion data and effectively improve advertising performance through specific improvement suggestions.

[1314] "Advertising performance data" refers to performance indicators such as click rates, conversion rates, and number of impressions that indicate the effectiveness of past advertising campaigns.

[1315] "Preprocessing" refers to the process of filling in missing values ​​in the data, filtering to maintain consistency, and normalizing the data to prepare it for analysis.

[1316] "User emotion data" refers to data that indicates the user's emotional state, obtained from the user's facial expression, voice tone, text input, and the like.

[1317] "AI analytics" is the process of using machine learning algorithms to analyze pre-processed data and evaluate the impact of each element of advertising creative on performance metrics and user sentiment.

[1318] "Points for improving advertising creative" refers to specific changes and improvements that are needed to improve advertising performance based on the results of AI analysis.

[1319] A "report" is a report that visually organizes the analysis results and extracted improvement points and displays them using charts, graphs, etc.

[1320] A "user's terminal" is a device such as a computer, smartphone, or tablet that a user uses to receive reports provided by the server.

[1321] This invention is a system aimed at efficiently improving advertising creatives, and provides a means for integrating and analyzing past advertising performance data and user emotion data. The system is primarily comprised of components: a server, a terminal, and a user. Here, we will explain in detail how this system is implemented.

[1322] Data collection

[1323] The server first sends an API request to an ad serving platform or database to collect past ad performance data. Specific APIs typically used are the Google Ads API and Facebook Ads API. This data includes click-through rates, conversion rates, and number of impressions.

[1324] The server also collects the user's emotional data using an emotion engine, such as Microsoft Azure Emotion API or IBM Watson Tone Analyzer, which analyzes and obtains the user's emotional state from their facial expressions, voice tone, and text input. Emotional data is collected when the device captures the user's video and audio and sends them to the server.

[1325] Data Preprocessing

[1326] The server preprocesses the advertising performance data and sentiment data collected by the server. The Python pandas library is used to impute missing values, and data filtering and normalization are also performed. For example, this includes outlier detection and normalization of the data scale using a scaling library.

[1327] AI analysis

[1328] The server then analyzes the pre-processed data using machine learning algorithms, specifically models such as random forests and neural networks. The analysis uses the scikit-learn library to train and evaluate the models. The AI ​​analysis evaluates the impact of each element of the ad creative on performance metrics and user sentiment.

[1329] Extracting points for improvement

[1330] The server extracts specific areas for improvement in the ad creative based on the results of the AI ​​analysis. It identifies common characteristics of ads with high click-through rates and conversion rates and summarizes the areas for improvement. For example, it generates suggestions for "bright colors" and "highly visible fonts."

[1331] Generate and deliver reports

[1332] The server visually organizes the extracted improvement points and generates a report using charts and graphs. Matplotlib and Plotly are used to generate the report, which is automatically created using a PDF generator. Finally, the server provides the generated report to the user's device. The report is delivered as a download link or email attachment, and the user can use it to design their next advertising campaign.

[1333] Specific examples

[1334] For example, suppose an e-commerce site runs a summer sale campaign ad with a 5% click-through rate, a 2% conversion rate, 100,000 impressions, and a 70% positive emotional response. In contrast, if a fall sale campaign ad has a 1% click-through rate, a 0.5% conversion rate, 80,000 impressions, and a 50% negative emotional response, the server collects and preprocesses this data and performs AI analysis. The analysis results suggest that bright colors and highly visible fonts are effective for summer ads.

[1335] The extracted improvement points are compiled into a report with specific suggestions, such as "Use brighter colors in the ad images and change the text font to a larger, more visible one." Finally, the server provides the generated report to the user's device, allowing the user to incorporate the improvements into their next ad campaign.

[1336] Prompt Sentence Examples

[1337] Please suggest areas for improvement in the ad creative based on the advertising data below.

[1338] (Successful summer sale campaign advertising data)

[1339] Click-through rate: 5%

[1340] Conversion rate: 2%

[1341] Views: 100,000

[1342] Positive emotional response: 70%

[1343] (Failed Fall Sale Campaign Advertising Data)

[1344] Click-through rate: 1%

[1345] Conversion rate: 0.5%

[1346] Views: 80,000

[1347] Negative emotional reactions: 50%

[1348] suggestion:

[1349] Use bright colors and highly visible fonts, just like in summer ads, to increase click-through rates and conversion rates.

[1350] With the above configuration, this system improves advertising performance and proposes advertising creatives that take user emotions into consideration.

[1351] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1352] Step 1: Collect advertising performance data

[1353] The server sends API requests to ad distribution platforms and databases to collect past ad performance data. Specifically, it uses the Google Ads API and Facebook Ads API. The input is API authentication information and the query required to obtain data, and the output is data on ad performance indicators such as click-through rate (CTR), conversion rate (CVR), and number of impressions. The server performs API authentication using the client ID and secret, and obtains data from the specified endpoint.

[1354] Step 2: Collecting emotion data

[1355] The server runs an emotion engine and collects emotion data from the user's facial expressions, voice tone, and text input. The inputs are camera footage, audio clips, and text data, and the emotional state is analyzed based on this. Specifically, the Microsoft Azure Emotion API and IBM Watson Tone Analyzer are used. The device captures the user's video and audio and sends them to the server to obtain emotion data. The output is data indicating the user's emotional state (e.g., smile rate, negative comment rate).

[1356] Step 3: Preprocessing the data

[1357] The advertising performance data and sentiment data collected by the server are preprocessed. The input is the raw data obtained in steps 1 and 2. Specific preprocessing includes missing value imputation, filtering, and normalization. The Python pandas library is used to impute missing values ​​with the mean or median, and outliers are filtered. Furthermore, a scaling library is used to normalize the data. The output is data that is ready for analysis.

[1358] Step 4: AI analysis

[1359] The server applies a machine learning algorithm to the preprocessed data and performs analysis. The input is the preprocessed advertising performance data and emotion data. Specifically, a random forest or neural network model is used, and the model is trained and evaluated using the scikit-learn library. The output is an evaluation of the impact of each element of the advertising creative on performance indicators and user emotions. The server randomly shuffles the training data and divides it into a training set and a test set. The algorithm is applied to train the model, and accuracy is verified on the test set.

[1360] Step 5: Identifying areas for improvement

[1361] The server extracts specific improvement points for the ad creative based on the results of the AI ​​analysis. The input is the data results of the AI ​​analysis. The server analyzes the importance of the features and displays a ranking of the elements that have the greatest impact on performance. The improvement points are written out as specific suggestions using natural language generation (NLG) technology. The output is specific changes and improvements (e.g., bright colors, highly visible fonts) as improvement suggestions.

[1362] Step 6: Generate reports

[1363] The server visually organizes the improvement points extracted and generates a report using charts and graphs. The input is improvement point data. Matplotlib and Plotly are used to generate the report. A data visualization library is used to graphically represent each indicator, and a PDF generator is used to automatically generate the report. The output is a report in PDF format or a web-based dashboard format.

[1364] Step 7: Providing the report

[1365] The server provides the generated report to the user's device. The input is the generated report data. The server configures the SMTP settings to send the report to the user's email address, and includes links and attachments in the email body. The output is a report provided to the user's device, which the user can use to design their next advertising campaign.

[1366] (Application example 2)

[1367] 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."

[1368] Conventional systems for improving advertising creatives rely solely on past performance data and are unable to consider the emotional responses of viewers. This makes it difficult to grasp the specific emotional impact that advertising has on viewers and to propose appropriate improvements based on that. In particular, there was a need for a method to evaluate the impact of visual and audio factors on viewer emotions, in addition to click-through rates and conversion rates.

[1369] 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.

[1370] In this invention, the server includes means for collecting performance data of past advertisements, means for collecting emotional response data, means for preprocessing the collected performance data and emotional response data, means for performing AI analysis based on the preprocessed data and emotional response data, means for extracting improvement points for the advertisement creative based on the results of the AI ​​analysis, means for generating a report including the improvement points, and means for providing the generated report to the user's connection device, thereby enabling efficient improvement of the advertisement creative based on the viewer's emotional response.

[1371] "Past advertising performance data" refers to data including performance indicators such as click rates, conversion rates, and number of impressions obtained in ad distribution.

[1372] "Preprocessing" refers to the process of preparing data suitable for analysis after collection by performing tasks such as filling in missing values, filtering, and normalizing the scale.

[1373] "Emotional response data" refers to emotional data of a user analyzed from facial expressions, tone of voice, text input, and the like while the user is viewing an advertisement.

[1374] "AI analysis" is the process of using machine learning algorithms to evaluate the impact of each element of advertising creative on performance metrics and emotions based on past advertising performance data and emotional response data.

[1375] "Ad creative" refers to the design elements of an ad, such as images, text, layout, and color usage.

[1376] "Improvement points" are specific elements or methods that are recommended to be changed or improved based on the results of AI analysis in order to improve the performance of advertising creatives.

[1377] "Report generation" is the process of visually organizing the AI ​​analysis results and points for improvement, and creating a report using charts and graphs to provide to the user.

[1378] "User's connected device" refers to the device (e.g., smartphone, tablet, PC) used by the user to provide the above-mentioned report.

[1379] A "machine learning algorithm" is an algorithm (e.g., random forest, neural network) that uses large amounts of data to train a model and make predictions or classifications from new data.

[1380] This invention is a system for efficiently improving advertising creatives, which utilizes a combination of past advertising performance data and user emotional response data. Specifically, it includes the following components:

[1381] First, the server uses an API to collect past advertising performance data from the ad distribution platform. This data includes performance indicators such as click-through rate, conversion rate, and number of impressions. Furthermore, a sentiment analysis engine is used to collect emotional responses from users' facial expressions, voice tone, text input, etc. This allows the server to also collect emotional data on users who view ads.

[1382] Next, the server preprocesses the collected ad performance data and emotional response data. Specifically, it imputes missing values, filters to maintain consistency, and normalizes the data scale. This prepares both data sets for analysis. Data preprocessing is performed using data processing libraries such as Pandas and Scikit-learn.

[1383] The server then uses machine learning algorithms to analyze the pre-processed ad performance data and emotional response data. At this stage, it uses tools such as random forests and neural networks (using TensorFlow and Keras) to create models that evaluate the impact of each element of the ad creative on performance indicators and emotions.

[1384] Based on the analysis results, the server can identify areas for improvement in ad creatives. For example, it may determine that ads with high click-through rates and conversion rates are effective with specific colors, fonts, and layouts, and that they also evoke positive emotional responses from users.

[1385] Points for improvement are compiled into specific proposals, which are then generated by the server as a report. The report is visually organized using charts and graphs and delivered to the user's connected device (smartphone, tablet, or PC). Visualization libraries such as Matplotlib and Seaborn are used to generate the report. Reports can also be created in PDF format using ReportLab.

[1386] As a specific example, consider a case where a server collects and analyzes the following data:

[1387] Summer sale advertisement:

[1388] Click-through rate: 5%

[1389] Conversion rate: 2%

[1390] Views: 100,000

[1391] Positive emotional responses (smiles, positive comments): 70%

[1392] Fall sale ad:

[1393] Click-through rate: 1%

[1394] Conversion rate: 0.5%

[1395] Views: 80,000

[1396] Negative emotional responses (sad facial expressions, negative comments): 50%

[1397] An example of a prompt sentence is as follows:

[1398] System Input:

[1399] The system collects data for a summer sale ad with a 5% click-through rate, a 2% conversion rate, 100,000 impressions, and a 70% positive emotional response, and a fall sale ad with a 1% click-through rate, a 0.5% conversion rate, 80,000 impressions, and a 50% negative emotional response.

[1400] Inference result:

[1401] We suggest that brightening the colors of advertising images and enlarging the font size would be effective.

[1402] This will enable efficient improvement of advertising creatives based on viewers' emotional responses.

[1403] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1404] Step 1:

[1405] The server uses the API from the ad distribution platform to collect data on past ad performance and user emotional response. This data includes click-through rates, conversion rates, number of impressions, users' facial expressions, voice tone, and text input. This data is the basis for measuring the effectiveness of ads and determining users' emotional responses.

[1406] Input: API of ad distribution platform, user facial expressions, voice tone, text input

[1407] Output: Collected advertising performance data and emotional response data

[1408] Step 2:

[1409] The ad performance data and emotional response data collected by the server are preprocessed. Specifically, missing values ​​are imputed, inconsistent data is filtered, and the data scale is normalized. Data processing libraries such as Pandas and Scikit-learn are used for preprocessing. Preprocessing improves data consistency and suitability for analysis.

[1410] Input: Collected advertising performance data and emotional response data

[1411] Output: Preprocessed ad performance data and emotional response data

[1412] Step 3:

[1413] The server performs AI analysis based on the preprocessed ad performance data and emotional response data. This analysis applies machine learning algorithms such as random forests and neural networks (using TensorFlow, Keras, etc.). The AI ​​analysis evaluates the impact of ad creative elements on performance indicators and user emotional responses and creates a model.

[1414] Input: Preprocessed advertising performance data and emotional response data

[1415] Output: A model for evaluating the impact of advertising creative elements on performance metrics and emotional responses.

[1416] Step 4:

[1417] Based on the results of the AI ​​analysis, the server extracts points for improvement in the ad creative. For example, it suggests specific ways to improve elements such as color usage, font, size, and placement. At this stage, feedback is obtained to suggest effective improvement measures based on the analysis results.

[1418] Input: A model for evaluating the impact of advertising creative elements on performance indicators and emotional responses.

[1419] Output: Specific suggestions for improvement

[1420] Step 5:

[1421] The server visually organizes the identified improvement points and generates a report, which can be created in PDF or web-based dashboard format and designed to include charts and graphs using visualization libraries such as Matplotlib, Seaborn, and ReportLab.

[1422] Input: Specific suggestions for improvement

[1423] Output: Visualized report

[1424] Step 6:

[1425] The server provides the generated report to the user's connected device (smartphone, tablet, PC), allowing the user to improve their ad creative and apply it to their next ad campaign to maximize the effectiveness of their ads.

[1426] Input: Visualized report

[1427] Output: Report provided to user's connected devices

[1428] 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.

[1429] 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.

[1430] 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.

[1431] 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.

[1432] 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.

[1433] 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.

[1434] 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).

[1435] 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.

[1436] 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."

[1437] 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.

[1438] 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).

[1439] 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.

[1440] 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.

[1441] 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.

[1442] 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.

[1443] 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.

[1444] 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.

[1445] 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.

[1446] 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.

[1447] 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.

[1448] 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.

[1449] The following is further disclosed regarding the above embodiment.

[1450] (Claim 1)

[1451] A means for collecting past advertising performance data;

[1452] means for preprocessing the collected data;

[1453] A means of performing AI analysis based on pre-processed data,

[1454] A means to extract improvements to advertising creatives based on the results of AI analysis,

[1455] means for generating a report including the improvement points;

[1456] The system includes means for providing the generated report to a user terminal.

[1457] (Claim 2)

[1458] 2. The system of claim 1, wherein the AI ​​analysis means uses a machine learning algorithm to evaluate the impact of advertising creative elements on performance indicators.

[1459] (Claim 3)

[1460] 2. The system according to claim 1, wherein the data preprocessing means performs missing value imputation, data filtering, and data normalization.

[1461] "Example 1"

[1462] (Claim 1)

[1463] A means for collecting past advertising performance data;

[1464] means for preprocessing the collected data;

[1465] A means of performing AI analysis based on pre-processed data,

[1466] A means to extract improvements to advertising creatives based on the results of AI analysis,

[1467] means for generating a report including the improvement points;

[1468] The system includes means for providing the generated report to a user terminal.

[1469] (Claim 2)

[1470] 2. The system of claim 1, wherein the AI ​​analysis means uses a data analysis algorithm to evaluate the impact of advertising creative elements on performance indicators.

[1471] (Claim 3)

[1472] 2. The system according to claim 1, wherein the data preprocessing means performs missing data complementation, data reconciliation, and data standardization.

[1473] "Application Example 1"

[1474] (Claim 1)

[1475] A means for collecting past advertising performance data;

[1476] means for preprocessing the collected data;

[1477] A means of performing AI analysis based on pre-processed data,

[1478] A means to extract improvements to advertising creatives based on the results of AI analysis,

[1479] means for generating a report including the improvement points;

[1480] means for providing the generated report to a user's terminal;

[1481] A means to provide real-time suggestions for improving advertising creatives,

[1482] The system includes means for generating a prompt sentence to respond to.

[1483] (Claim 2)

[1484] 2. The system of claim 1, wherein the AI ​​analysis means uses a machine learning algorithm to evaluate the impact of advertising creative elements on performance indicators.

[1485] (Claim 3)

[1486] 2. The system according to claim 1, wherein the data preprocessing means performs missing value imputation, data filtering, and data normalization.

[1487] "Example 2: Combining Emotion Engines"

[1488] (Claim 1)

[1489] A means for collecting past advertising performance data;

[1490] means for preprocessing the collected data;

[1491] means for collecting user emotion data;

[1492] A means of performing AI analysis based on pre-processed data,

[1493] A means to extract improvements to advertising creatives based on the results of AI analysis,

[1494] means for generating a report including the improvement points;

[1495] The system includes means for providing the generated report to a user terminal.

[1496] (Claim 2)

[1497] 2. The system of claim 1, wherein the AI ​​analysis means uses machine learning algorithms to evaluate the impact of advertising creative elements on performance indicators and user emotions.

[1498] (Claim 3)

[1499] 2. The system according to claim 1, wherein the data preprocessing means performs missing value imputation, data filtering, and data normalization.

[1500] "Application example 2 when combining emotion engines"

[1501] (Claim 1)

[1502] A means for collecting past advertising performance data;

[1503] means for preprocessing the collected data;

[1504] A means of conducting AI analysis based on pre-processed data and emotional response data;

[1505] A means to extract improvements to advertising creatives based on the results of AI analysis,

[1506] means for generating a report including the improvement points;

[1507] The system includes means for providing the generated report to a user's connected device.

[1508] (Claim 2)

[1509] 2. The system of claim 1, wherein the AI ​​analysis means uses machine learning algorithms to evaluate the impact of advertising creative elements on performance metrics and emotional responses.

[1510] (Claim 3)

[1511] 2. The system according to claim 1, wherein the data preprocessing means performs missing value imputation, data filtering, and data normalization. [Explanation of symbols]

[1512] 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 past advertising performance data; means for preprocessing the collected data; A means of performing AI analysis based on pre-processed data, A means to extract improvements to advertising creatives based on the results of AI analysis, means for generating a report including the improvement points; The system includes means for providing the generated report to a user terminal.

2. 2. The system according to claim 1, wherein the AI ​​analysis means uses a machine learning algorithm to evaluate the impact of advertising creative elements on performance indicators.

3. 2. The system according to claim 1, wherein the data preprocessing means performs missing value imputation, data filtering, and data normalization.

Images