System

Abstract

A system is provided.SOLUTION: A system, comprising: means for collecting sales data, market data, store data, and salesperson data; means for pre-processing the collected data; means for analyzing the pre-processed data and extracting features; means for generating a role play scenario based on the extracted features; means for distributing the generated scenario to a terminal of each salesperson; means for performing role play by the salesperson based on the scenario and collecting feedback; and means for updating a machine learning model based on the collected feedback.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 role-play scenarios in sales training often contain general stereotypical content and do not adequately address the characteristics of individual salespeople or stores. This limits the effectiveness of the training and makes it difficult to translate into improved performance in actual sales situations. To solve this problem, it is necessary to provide customized role-play scenarios based on the characteristics of each salesperson and store. [Means for solving the problem]

[0005] The system of the present invention includes means for collecting sales data, market data, store data, and salesperson data, preprocessing the data, and then analyzing the data to extract features. This allows for an accurate understanding of the characteristics of salespersons and stores. Furthermore, by introducing means for generating role-play scenarios based on the extracted features, it provides scenarios optimized for individual salespersons and stores. These scenarios are distributed to the salesperson's terminal, and feedback on the role-plays performed by the salesperson is collected. The machine learning model is updated based on the collected feedback and reflected in the next scenario generation. This series of processes maximizes the effectiveness of training and improves salesperson performance.

[0006] "Sales data" refers to sales information, transaction data, customer information, etc. for specific products and services obtained during sales activities.

[0007] "Market data" refers to various statistical information related to the sales environment, such as economic conditions, market trends, and the status of competitors.

[0008] "Store data" is information about a specific store, including, for example, the location, number of customers, sales data, and the like.

[0009] "Salesperson data" is information about each salesperson, including, for example, past sales performance, skills, training history, and the like.

[0010] "Preprocessing" refers to the process of removing duplicates and missing data from collected data and preparing it in a format that is easy to analyze.

[0011] "Features" refer to patterns and trends extracted from analyzed data, which indicate the characteristics of salespeople and stores.

[0012] A "role-play scenario" refers to a scenario that simulates interactions with fictitious customers, designed to help trainees acquire practical skills during sales training.

[0013] "Device" means an electronic device used to conduct role-play scenarios and collect and send feedback, including, for example, a tablet, computer, or smartphone.

[0014] "Feedback" refers to information such as comments, evaluations, and areas for improvement provided by the salesperson after the role-play.

[0015] A "machine learning model" refers to an algorithm or statistical model that automatically learns from data and efficiently performs specific tasks. [Brief explanation of the drawings]

[0016] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11]FIG. 3 is a sequence diagram 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

[0017] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

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

[0019] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0020] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0021] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0022] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0023] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0024] [First embodiment]

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

[0026] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0027] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0028] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0029] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0030] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0031] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

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

[0033] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0034] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0035] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0036] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0037] The system according to the present invention has the function of collecting sales data, market data, store data, and salesperson data, and then generating, distributing, and evaluating role-play scenarios based on this data. The specific processing of this program and a natural language explanation are given below.

[0038] First, the server collects data from external and internal sources, including store sales data, foot traffic, market data, and each salesperson's past performance data, and stores the collected data in a database.

[0039] The server then preprocesses the stored data, which includes cleansing the data, standardizing the format, and removing duplicate data. For example, it imputes missing data and removes irrelevant data. At this stage, the preprocessed data is prepared for analysis.

[0040] After the preprocessing is complete, the server performs data analysis. This analysis is carried out using statistical methods and machine learning algorithms to extract data characteristics. For example, analyzing the data of Salesperson A reveals that he tends to be strong at certain times of the day and weak at other times of the day.

[0041] Next, the server generates role-play scenarios optimized for each salesperson based on the extracted features. This scenario generation uses neural networks and other AI techniques. For example, a scenario is generated for Salesperson A to improve his or her closing skills with customers during weak times.

[0042] The generated scenario is distributed by the server to the terminal of the target salesperson. Once the scenario is distributed, the user (salesperson) uses the terminal to perform role-playing. After the role-playing is completed, the user enters feedback into the terminal. The feedback includes the effectiveness of the scenario, areas for improvement, and performance.

[0043] The collected feedback is sent to the server. The server analyzes the feedback data and updates the machine learning model. This update improves the accuracy of the next scenario generation. For example, by incorporating feedback from Salesperson A, a more effective next scenario is generated.

[0044] In this way, the system provides each salesperson with a customized role-play scenario, maximizing the effectiveness of the training. This system is expected to improve salesperson performance.

[0045] The processing flow will be explained below.

[0046] Specific processing flow of the program

[0047] Step 1:

[0048] The server collects data from external and internal sources, specifically sales data, market data, store data, and salesperson data, using APIs and database queries. For example, it retrieves sales data by store, market trend data, and salesperson historical performance data.

[0049] Step 2:

[0050] The server preprocesses the collected data. Specifically, it complements missing data, removes duplicate data, and standardizes the format. For example, it complements missing data with average values ​​and removes inappropriate data.

[0051] Step 3:

[0052] The server analyzes the preprocessed data. Specifically, it uses statistical techniques and machine learning models to extract patterns and features from the data. For example, it can discover whether there are strong or weak trends on certain days of the week or during certain times of the day from the data of Salesperson A.

[0053] Step 4:

[0054] The server generates role-play scenarios based on the extracted characteristics. Specifically, it uses neural networks and AI technology to generate scenarios tailored to the characteristics of each salesperson and store. For example, for Salesperson A, it creates a scenario that includes closing techniques that should be strengthened during times when Salesperson A is weak.

[0055] Step 5:

[0056] The server distributes the generated role-play scenario to the device of the salesperson. Specifically, the scenario is sent to the device using notifications or file transfers. For example, the scenario is pushed to the tablet of Salesperson A.

[0057] Step 6:

[0058] The user (salesperson) uses the terminal to carry out a role-play scenario. Specifically, the user reads the displayed scenario and acts based on it. For example, Salesperson A performs a role-play according to the scenario.

[0059] Step 7:

[0060] The user enters feedback on the role-play into the terminal. Specifically, the results and impressions are entered into a form on the terminal and sent. For example, Salesperson A enters feedback such as "The closing technique in the scenario was effective."

[0061] Step 8:

[0062] The server collects the feedback and stores it in a database. Specifically, it receives and stores the feedback data sent from the terminal. For example, the feedback of salesperson A is stored in the database.

[0063] Step 9:

[0064] The server analyzes the collected feedback and updates the machine learning model. Specifically, it learns from new data and improves the model's accuracy. For example, the model is retrained using feedback data from Salesperson A.

[0065] Through these steps, a series of processes are implemented to provide each salesperson with the most appropriate role-play scenario, thereby maximizing the effectiveness of sales training.

[0066] Example 1

[0067] 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."

[0068] Conventional salesperson training systems have the problem of being dependent on a uniform training program, making it difficult to customize based on the characteristics and performance of each individual salesperson. This means that it is difficult to effectively improve the performance of each salesperson, and the effectiveness of training is limited. In addition, there is a lack of a mechanism for flexibly improving the training program based on feedback, which creates challenges in long-term salesperson development.

[0069] 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.

[0070] In this invention, the server includes means for collecting sales data, market data, store data, and salesperson data, means for preprocessing the collected data, means for analyzing the preprocessed data using statistical techniques and machine learning algorithms to extract features, means for generating role-play scenarios using a neural network or the like based on the extracted features, means for distributing the generated scenarios to the terminals of each salesperson, means for the salespersons to use their terminals to perform role-plays and collect feedback, and means for analyzing the collected feedback and updating the machine learning model. This makes it possible to provide a training program customized for each salesperson and enable continuous improvement based on the feedback.

[0071] "Sales Data" refers to information relating to the sales of goods and services at a store.

[0072] "Market Data" means information about external market trends and economic factors.

[0073] "Store data" refers to information regarding the number of customers visiting the store and the store's operating status.

[0074] "Salesperson Data" means information about the accomplishments and performance of individual salespersons.

[0075] "Preprocessing" refers to processing the collected data such as filling in missing values, standardizing the format, and deleting inappropriate data.

[0076] "Analysis" refers to the extraction of features and patterns from data using statistical methods and machine learning algorithms.

[0077] A "role play scenario" is a training scenario used by salespeople for practice or simulation.

[0078] A "neural network" is a type of algorithm consisting of multiple layers of artificial neurons and used for pattern recognition and machine learning.

[0079] A "terminal" is a device, such as a computer or tablet, used by a salesperson.

[0080] "Feedback" refers to opinions and results regarding the effectiveness of the scenario and areas for improvement that the salesperson enters after performing the role-play scenario.

[0081] A "machine learning model" is an algorithm that analyzes data and makes predictions and classifications based on the results.

[0082] The system of the present invention optimizes salesperson training through multiple steps: data collection, preprocessing, analysis, scenario generation, scenario distribution, role-playing, feedback collection, feedback analysis, and updating of machine learning models. This system provides role-playing scenarios customized for each salesperson, enabling effective training.

[0083] First, the server collects sales data, market data, store data, and salesperson data. Specifically, sales data is obtained from the POS system, market data from an external API, store data from customer count sensors, and salesperson data from the HR system. These data are integrated and stored in a database on the server.

[0084] The server then pre-processes the collected data, which includes data cleansing (filling in missing data, removing irrelevant data), standardizing formats (for example, standardizing date formats), and removing duplicate data, making the data suitable for analysis.

[0085] Next, the server analyzes the preprocessed data. Statistical methods and machine learning algorithms are used for the analysis. Specific tools used include Python's Pandas and Scikit-learn. At this stage, data features and patterns are extracted. For example, analyzing the sales data of Salesperson A may reveal a strong sales trend during certain time periods.

[0086] After the features are extracted, the server uses a generative AI model such as a neural network to generate role-play scenarios based on these features. The scenarios include content that addresses specific weaknesses of the salesperson. For example, a scenario is generated for Salesperson A to improve his or her closing skills with customers during weak times.

[0087] The generated scenario is sent from the server to the salesperson's terminal. After the scenario is sent, the user (salesperson) uses the terminal to perform the role-play. Examples of specific prompts for the role-play include the following:

[0088] Prompt: "How can I approach a male customer in his 30s who visits my store between 3:00 PM and 5:00 PM on Sundays?"

[0089] After the user performs a role-play based on the scenario, they input feedback into the terminal, including the effectiveness of the scenario, areas for improvement, and performance.

[0090] The collected feedback is sent to a server, which analyzes it and updates the machine learning model based on the analysis results. This update improves the accuracy of the next scenario generation, making training for salespeople even more effective.

[0091] This system provides training scenarios optimized for each salesperson, maximizing the effectiveness of the training. For example, by implementing the scenarios provided by the system, Salesperson A's success rate in closing sales improved even during times when he was previously weak at dealing with customers.

[0092] In this way, the present invention improves the accuracy and effectiveness of salesperson training through a series of processes from data collection to model updating.

[0093] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0094] Step 1: Data collection

[0095] The server collects sales data, market data, store data, and salesperson data. Data collection is performed using API requests to obtain data from the POS system, external APIs, customer count sensors, and HR systems. Each piece of data is sent to the server in JSON or CSV format. The specific operation of data collection is for the server to periodically make API calls from each data source and store the obtained data in a database.

[0096] Input: Various external and internal data obtained by API requests

[0097] Output: Sales data, market data, store data, salesperson data stored in the database

[0098] Step 2: Data Preprocessing

[0099] The server preprocesses the collected data. Preprocessing includes data cleansing, format standardization, and removal of duplicate data. Missing data is filled in with the average value from the past, and inappropriate data is removed. To standardize the format, the date format is standardized to "YYYY-MM-DD." Specifically, the server uses the Python Pandas library to manipulate data frames and perform cleansing and format standardization.

[0100] Input: Raw data stored in the database (sales data, market data, store data, salesperson data)

[0101] Output: Preprocessed data (cleaned data)

[0102] Step 3: Data analysis

[0103] The server analyzes the preprocessed data. Statistical methods and machine learning algorithms are used for this analysis. Specific methods include regression analysis and clustering using Python's Scikit-learn. The server extracts each salesperson's characteristics and trends based on the salesperson data. For example, it can discover the sales patterns of salesperson A, who is strong during specific time periods.

[0104] Input: Preprocessed data (cleaned data)

[0105] Output: Analyzed feature data (statistical results and machine learning model output)

[0106] Step 4: Scenario generation

[0107] The server generates role-play scenarios based on the analyzed feature data. Using neural networks and generative AI models, optimal scenarios are created for each salesperson. For example, a scenario is generated for Salesperson A to improve his closing skills with customers during a specific time period. The generated scenarios are converted into natural-sounding sentences using natural language generation (NLG) technology.

[0108] Input: Analyzed feature data (features for each salesperson)

[0109] Output: Generated role-play scenario (text format)

[0110] Step 5: Scenario distribution and execution

[0111] The server distributes the generated scenario to each salesperson's device. The salesperson (user) uses the device to execute the scenario. For example, Salesperson A performs a customer service simulation while reading the prompts displayed on the device. Push notifications and APIs are used to distribute the scenario.

[0112] Input: Generated role-play scenario (text format)

[0113] Output: Scenario delivered to the terminal (text displayed on the salesperson's terminal)

[0114] Step 6: Gather feedback

[0115] After the user (salesperson) performs the role-play scenario, they input their feedback into the terminal. The feedback includes the effectiveness of the scenario, areas for improvement, and performance. The terminal then sends this feedback data to the server.

[0116] Input: Feedback entered by the salesperson into the terminal (text format)

[0117] Output: Feedback data sent to the server (JSON or CSV format)

[0118] Step 7: Analyze feedback and update the model

[0119] The server analyzes the collected feedback data and updates the machine learning model. For analysis, the server uses a Python library to retrain the model parameters based on the feedback data, which makes the next scenario generation more accurate.

[0120] Input: Feedback data stored on the server (text, JSON, CSV)

[0121] Output: Updated machine learning model (improved AI model)

[0122] This series of processes systematically supports salespeople in improving their performance, and continuous learning effects can be expected.

[0123] (Application example 1)

[0124] 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."

[0125] Aiming to improve the performance of salespeople in brick-and-mortar stores, there is a need to effectively improve sales skills, especially closing skills, by providing training scenarios tailored to each individual salesperson. Traditional training methods have struggled to effectively incorporate feedback and ensure continuous improvement. Furthermore, there are limited means of evaluating training content in real time using technologies such as voice recognition. Therefore, a system that can accurately address each salesperson's weaknesses is needed.

[0126] 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.

[0127] In this invention, the server includes means for collecting sales data, market data, store data, and salesperson data, means for preprocessing the collected data, means for analyzing the preprocessed data and extracting features, means for generating role-play scenarios based on the extracted features, means for distributing the generated scenarios to terminals of each salesperson, means for the salesperson to perform role-plays based on the scenarios and collect feedback, means for updating a machine learning model based on the collected feedback, means for generating scenarios for the salesperson to improve their closing skills for a specific product, means for the terminal to automatically evaluate responses in the role-plays using a voice recognition function, and means for using the feedback data to reflect the feedback data in the generation of the next scenario. This makes it possible to generate and evaluate training scenarios tailored to individual salespersons and reflect the feedback.

[0128] "Sales data" is information relating to product sales, including sales quantity, sales amount, sales date, sales time, and the like.

[0129] "Market data" is information about the current market situation, such as competitor trends, fluctuations in product prices, and consumer purchasing intentions.

[0130] "Store data" is information about a specific store, including the number of customers, inventory status, store layout, business hours, and the like.

[0131] "Salesperson data" is information about each salesperson, including past sales performance, customer service skills, working hours, and training history.

[0132] "Preprocessing" refers to preparing the collected raw data in a format that can be analyzed, and includes steps such as data cleansing, format standardization, and missing value imputation.

[0133] "Feature extraction" refers to finding important patterns and trends through data analysis, using statistical analysis and machine learning algorithms to reveal the essential parts of the data.

[0134] A "role-play scenario" is a training scenario that allows salespeople to improve their skills by interacting with virtual customers, and includes specific customer service situations and troubleshooting methods.

[0135] "Terminal" refers to the device used by the salesperson, which may be a smartphone, tablet, or other portable computing device.

[0136] "Feedback" refers to salespeople providing their opinions on the effectiveness of the scenario and areas for improvement after the role-play, and this information is collected and used as valuable data to improve the system.

[0137] A "machine learning model" is an algorithm that generates patterns and predictions based on past data. It is used to update the model based on collected feedback and improve the quality of training scenarios.

[0138] The "voice recognition function" is a technology that allows the terminal to analyze the salesperson's voice in real time and process the spoken content as text data, and is used to automatically evaluate the responses in role-playing.

[0139] "Closing skills" are techniques for getting customers to make a purchase decision at the final stage of sales negotiations or customer service, and are important abilities that include persuasiveness and communication skills.

[0140] The system according to the present invention has the function of collecting sales data, market data, store data, and salesperson data, and then generating, distributing, and evaluating role-play scenarios based on these data. The specific processing of this system is explained below in natural language.

[0141] First, the server collects data from external and internal sources, including store sales data, foot traffic, market data, and each salesperson's past performance data, and stores the collected data in a database.

[0142] The server then preprocesses the stored data, which includes cleansing the data, standardizing the format, and removing duplicate data. For example, it imputes missing data and removes irrelevant data. At this stage, the preprocessed data is prepared for analysis.

[0143] After the preprocessing is complete, the server performs data analysis. This analysis is carried out using statistical methods and machine learning algorithms to extract data characteristics. For example, analyzing the data of Salesperson A reveals that he tends to be strong at certain times of the day and weak at other times of the day.

[0144] Next, the server generates role-play scenarios optimized for each salesperson based on the extracted features. Neural networks and other AI techniques are used to generate these scenarios. For example, a scenario is generated for Salesperson A to improve his closing skills with customers during weak times. The prompt might be something like, "Salesperson A is having trouble closing sales for a specific product. Please generate a role-play scenario to improve his skills in promoting this product."

[0145] The generated scenario is distributed by the server to the target salesperson's terminal. Once the scenario is distributed, the salesperson uses the terminal to perform role-playing. For example, Salesperson A receives a scenario to practice closing a specific product, and then interacts with a virtual customer based on that scenario.

[0146] After the role-play, the salesperson enters feedback into the terminal, including the effectiveness of the scenario, areas for improvement, and performance. The terminal may also automatically evaluate the responses during the role-play using voice recognition.

[0147] The collected feedback is sent to the server. The server analyzes the feedback data and updates the machine learning model. This update improves the accuracy of the next scenario generation. For example, by incorporating feedback from Salesperson A, a more effective next scenario is generated.

[0148] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0149] Step 1:

[0150] First, the server collects data from external and internal sources, including store sales data, customer traffic, market data, and each salesperson's past performance data. The input is data from the sales system, customer management system, and market information system, and the output is stored in a database.

[0151] Step 2:

[0152] Next, the server preprocesses the stored data. This includes data cleansing, formatting standardization, and deleting duplicate data. The input is the raw data collected in step 1, and the output is data in a clean, analyzable format. Specifically, missing values ​​are imputed, inappropriate data is deleted, and the format is standardized.

[0153] Step 3:

[0154] The server then performs data analysis based on the preprocessed data. Statistical methods and machine learning algorithms are used for the analysis. For example, based on the data of Salesperson A, strengths and weaknesses in specific time periods can be extracted. The input is the preprocessed data, and the output is an index showing the characteristics of the data.

[0155] Step 4:

[0156] Next, the server generates role-play scenarios optimized for each salesperson based on the extracted features. Neural networks and other AI techniques are used to generate the scenarios. For example, a prompt such as "Salesperson A is struggling to close a specific product. Please generate a role-play scenario to help him improve his skills in promoting this product" can be input, generating a scenario focused on specific closing skills. The output is a specific role-play scenario.

[0157] Step 5:

[0158] The generated scenario is delivered by the server to the target salesperson's terminal. The input is the generated role-play scenario, and the output is the delivered scenario. The salesperson can receive the role-play scenario on their terminal.

[0159] Step 6:

[0160] After the scenario is delivered, the salesperson uses the terminal to perform the role-play. Following the scenario on the terminal, the salesperson interacts with a virtual customer. The input is the delivered scenario, and the output is the result of the role-play.

[0161] Step 7:

[0162] After the role-play, the salesperson enters their feedback into the terminal. The feedback includes the effectiveness of the scenario, areas for improvement, and performance. The input is the results of the role-play and the salesperson's opinions, and the output is structured feedback data.

[0163] Step 8:

[0164] The collected feedback is sent to the server, which analyzes the feedback data and updates the machine learning model. The input is the feedback data, and the output is an updated machine learning model. This improves the accuracy of the next scenario generation.

[0165] Through these steps, a series of processes are carried out, providing each salesperson with a role-play scenario that is optimized for them.

[0166] 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.

[0167] This invention relates to a system that collects and analyzes sales data, market data, store data, and salesperson data, and generates role-play scenarios based on these data. Furthermore, by combining this with an emotion engine that recognizes the emotions of users (salespersons), more effective training can be achieved.

[0168] First, the server collects sales data, market data, store data, and salesperson data from external and internal sources, using APIs and database queries to retrieve various data and store it in a database.

[0169] Next, the server preprocesses the collected data. Since the collected data is often incomplete, missing data is filled in, duplicate data is removed, and the format is standardized. This prepares the data for analysis.

[0170] The server then performs data analysis based on the preprocessed data. Statistical methods and machine learning models are used to extract the characteristics of each salesperson and store. For example, patterns can be discovered, such as strengths on certain days of the week or at certain times of the day, but weaknesses on other days or at certain times of the day.

[0171] The server then generates role-play scenarios based on the extracted features. Using neural networks and other AI techniques, scenarios optimized for each salesperson and store are generated. For example, for Salesperson A, a scenario is created to strengthen closing skills to overcome weak performance.

[0172] The generated scenario is delivered from the server to the target salesperson's terminal. The scenario is sent to the terminal using notifications or file transfers. The user (salesperson) then performs a role-play based on the delivered scenario.

[0173] When the user is performing a role-play, the emotion engine installed in the device analyzes the salesperson's facial expressions and tone of voice to recognize their emotions. For example, if Salesperson A is feeling stressed, this information will be detected by the emotion engine.

[0174] The emotional information analyzed by the emotion engine is sent to the server. The server then dynamically adjusts the difficulty and content of the role-play scenario based on this information. For example, if Salesperson A feels stressed by a difficult scenario, the difficulty of the scenario can be lowered.

[0175] After the role-playing is completed, the user inputs feedback into the terminal. The feedback includes the results of the role-playing, impressions, and areas for improvement. The server collects this feedback and stores it in a database.

[0176] The collected feedback is then analyzed by the server again to update the machine learning model, which will enable it to provide more accurate and effective scenarios the next time it generates them.

[0177] In this way, by providing customized role-play scenarios that take into account the emotions of salespeople, it is expected that the effectiveness of training will be maximized and salespeople's performance will improve.

[0178] The processing flow will be explained below.

[0179] MODE FOR CARRYING OUT THE INVENTION

[0180] Step 1:

[0181] The server collects sales data, market data, store data, and salesperson data from external and internal sources. Specifically, it retrieves various data using APIs and database queries and stores it in a database. For example, store sales data and market trend data can be retrieved through an API and stored in the appropriate database tables.

[0182] Step 2:

[0183] The server preprocesses the collected data. Specific operations include filling in missing data, deleting duplicate data, and standardizing formats. For example, if there are gaps in sales data, it fills in the gaps with the average value from the past, and if the data format is inconsistent, it converts it into a unified format.

[0184] Step 3:

[0185] The server then performs data analysis using the preprocessed data. Specifically, statistical methods and machine learning models are used to extract patterns and features from the data. For example, analysis can be performed to discover a trend in which Salesperson A's performance is weak during certain time periods.

[0186] Step 4:

[0187] The server generates role-play scenarios based on the extracted features. As a specific example, neural networks and AI technology are used to create scenarios that strengthen salespeople's weaknesses. For example, a role-play scenario is generated to strengthen Salesperson A's closing skills when dealing with customers during weekdays.

[0188] Step 5:

[0189] The server generates a role-play scenario and distributes it to the salesperson's device. Specifically, it uses notification and file transfer functions to send the scenario to the salesperson's tablet or smartphone.

[0190] Step 6:

[0191] The user (salesperson) uses the terminal to perform role-playing. Specifically, the user follows a scenario displayed on the terminal to simulate customer interactions. For example, a role-playing session is conducted in which Salesperson A tries out a new closing technique.

[0192] Step 7:

[0193] During the role-play, the device's built-in emotion engine recognizes the user's emotions. For example, it uses a camera and microphone to analyze the salesperson's facial expressions and tone of voice, collecting emotional data such as stress and satisfaction.

[0194] Step 8:

[0195] The server receives the emotion data sent from the emotion engine and dynamically adjusts the difficulty and content of the role-play scenario. For example, if Salesperson A is feeling stressed, the server adjusts the difficulty of the next scenario based on this information.

[0196] Step 9:

[0197] The user enters feedback on the role-play into the terminal. Specifically, after the role-play is completed, the user enters the results, impressions, and points for improvement into the feedback form and submits it. For example, Salesperson A enters feedback such as, "The closing techniques in the scenario were effective."

[0198] Step 10:

[0199] The server analyzes the collected feedback and stores it in a database. Specifically, it receives and analyzes the feedback data to update the machine learning model. This data is reflected in the next scenario generation. For example, the feedback from Salesperson A is used to learn how to create a more precise scenario for the next time.

[0200] This process maximizes the effectiveness of training by providing each salesperson with tailored role-play scenarios.

[0201] Example 2

[0202] 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."

[0203] Conventional salesperson training systems have difficulty providing real-time feedback on salespersons' actual performance and emotions, making it impossible to provide training scenarios optimized for each individual salesperson. Furthermore, there was no system in place to properly collect and analyze training feedback and reflect it in the next training session. As a result, the effectiveness of training was limited.

[0204] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0205] In this invention, the server includes means for collecting sales data, market data, store data, and salesperson data, means for preprocessing the collected data, means for analyzing the preprocessed data and extracting features, means for generating role-play scenarios based on the extracted features using a neural network or other generative AI model, means for distributing the generated scenarios to each salesperson's terminal, means for the salesperson to perform role-plays based on the scenarios and for recognizing emotions using an emotion engine of the terminal, means for dynamically adjusting the difficulty and content of the role-play scenarios based on the emotion information, means for collecting feedback after the scenarios are performed, and means for updating the machine learning model based on the collected feedback. This makes it possible to provide effective training scenarios based on the salesperson's actual performance and emotions and to reflect the feedback in real time.

[0206] "Sales Data" refers to information related to the sale of goods or services, including sales quantity, sales amount, sales date and time, customer information, etc.

[0207] "Market data" refers to information related to market conditions, including the balance of supply and demand, price fluctuations, competitive conditions, and economic indicators.

[0208] "Store data" refers to information related to a specific store, including the store's location, size, business hours, staff information, sales performance, etc.

[0209] "Salesperson Data" means information related to individual salespersons, including their skills, experience, ratings, and performance.

[0210] "Preprocessing" refers to tasks such as data cleaning and format conversion to convert collected data into an analyzable format.

[0211] "Feature extraction" refers to the process of finding important patterns and relationships in pre-processed data.

[0212] "Generative AI models" refer to technologies that use neural networks and other artificial intelligence algorithms to generate new information and scenarios from data.

[0213] A "role-play scenario" refers to a scenario created for the purpose of training salespeople to practice sales activities in specific situations.

[0214] "Terminal" refers to the digital device used by the salesperson, including PCs, tablets, smartphones, etc.

[0215] An "emotion engine" refers to software or hardware that analyzes and recognizes emotions from a salesperson's facial expressions, tone of voice, etc.

[0216] "Feedback" refers to comments and evaluations provided by salespeople after the role-play, including information on the effectiveness of the training and areas for improvement.

[0217] A "machine learning model" is an algorithm that makes predictions or classifications based on data and is trained using training data.

[0218] The present invention relates to a system that collects and analyzes sales data, market data, store data, and salesperson data, and generates role-play scenarios based on these data. Furthermore, by combining this system with an emotion engine that recognizes the emotions of users (salespersons), it is possible to achieve more effective training. Specific embodiments are described below.

[0219] First, the server collects sales data, market data, store data, and salesperson data from external and internal sources. To do this, it uses APIs and database queries to retrieve various data and stores it in a database. Specifically, it uses RESTful APIs and SQL queries. For example, the server retrieves market data from "https: / / api.example.com / marketdata" and retrieves sales data with the SQL query "SELECT FROM sales WHERE date='2023-10-01'".

[0220] Next, the data collected by the server is preprocessed. As the data is often incomplete as it is, missing data is complemented, duplicate data is deleted, and the format is standardized. Specifically, using the Python Pandas library, duplicate data is deleted using the ".drop_duplicates()" method, and missing data is complemented using the ".fillna(method='ffill')" method.

[0221] The server performs data analysis based on the preprocessed data. It uses statistical methods and machine learning models (e.g., regression analysis, clustering) to extract the characteristics of each salesperson and store. For example, the server uses the Scikit-learn library to analyze the relationship between sales data and days of the week using the "LinearRegression()" model.

[0222] Next, the server generates role-play scenarios based on the analysis results. It uses neural networks or generative AI models (e.g., GPT-3 (registered trademark), BERT) to generate scenarios optimized for each salesperson and store. Specifically, it uses the TENSORFLOW (registered trademark) library to create a neural network model with "tf.keras.models.Sequential()" to generate a scenario specifically for Salesperson A.

[0223] The generated scenario is distributed from the server to the terminal of the target salesperson. The scenario is sent to the terminal using notification or file transfer. Specifically, the file is transferred using an HTTP request or FTP. For example, the server uses "requests.post(url, data=scenario data)" to send the scenario to Salesperson A's terminal.

[0224] The user (salesperson) performs a role-play based on the delivered scenario. During the role-play, an emotion engine (e.g., OpenFace or IBM Watson (registered trademark) Tone Analyzer) installed in the device analyzes the salesperson's facial expressions and tone of voice to recognize their emotions. For example, the device uses the OpenFace library to analyze emotions such as "stress" or "confidence" from the user's facial expression data.

[0225] The emotional information analyzed by the emotion engine is sent to the server. The server uses this information to dynamically adjust the difficulty and content of the role-play scenario. Specifically, it uses a reinforcement learning model to change the difficulty of the scenario in real time. For example, if it receives information that Salesperson A is feeling stressed, it will set the difficulty of the scenario to a lower level.

[0226] After the role-play is completed, the user enters feedback into the device. The feedback includes the results of the role-play, impressions, and areas for improvement. Specifically, opinions are collected using Google (registered trademark) Forms or a feedback form developed in-house. For example, the user may enter their impression that "my closing skills have improved" into the form displayed on the device.

[0227] The server analyzes the collected feedback and updates the machine learning model. The feedback data is used to improve the model's accuracy and is reflected in the next scenario generation. Specifically, the feedback data is cleaned using the Pandas library and the model is retrained using Scikit-learn.

[0228] Prompt Sentence Examples

[0229] "You are the server of a system that collects and analyzes sales data, market data, store data, and salesperson data, and generates role-play scenarios based on this data. Specifically, please perform the following tasks:

[0230] 1. Collect data using API queries or SQL queries and store it in a database.

[0231] 2. Preprocess the data using the Pandas library.

[0232] 3. Analyze the data using a machine learning model to extract the characteristics of each salesperson.

[0233] 4. Generate role-play scenarios using TensorFlow.

[0234] 5. The scenario is delivered to the user's device via an HTTP request.

[0235] 6. The user performs a role-play based on the scenario, and the device's emotion engine recognizes emotions.

[0236] 7. Adjust the difficulty of the scenario based on emotional data.

[0237] 8. Collect user feedback and incorporate it into the next scenario generation.

[0238] For example, if Salesperson A is feeling stressed, adjust the scenario to be less difficult and use the feedback data to update the machine learning model for effective training.

[0239] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0240] Step 1: Data collection

[0241] The server obtains sales data, market data, store data, and salesperson data from external and internal sources.

[0242] Input: API query, SQL query

[0243] Output: Collected data (sales data, market data, store data, salesperson data)

[0244] The server retrieves market information using a RESTful API and pulls store information from an internal database using SQL queries. For example, it retrieves market data from "https: / / api.example.com / marketdata" and retrieves sales data with the SQL query "SELECT FROM sales WHERE date='2023-10-01'".

[0245] Step 2: Data Preprocessing

[0246] The server preprocesses the collected data.

[0247] Input: Collected data

[0248] Output: Preprocessed data

[0249] The server uses the Python Pandas library to clean the data. Specifically, it removes duplicate data using the ".drop_duplicates()" method, fills missing data using the ".fillna(method='ffill')" method, and standardizes the data format.

[0250] Step 3: Data analysis

[0251] The server performs analysis based on the preprocessed data and extracts features.

[0252] Input: Preprocessed data

[0253] Output: Extracted feature data

[0254] The server uses the Scikit-learn library to analyze the data using statistical methods and machine learning models (e.g., regression analysis, clustering). Specifically, it uses the "LinearRegression()" model to analyze the relationship between sales data and days of the week and extract features.

[0255] Step 4: Role-play scenario generation

[0256] The server generates a role-play scenario based on the extracted features.

[0257] Input: Extracted feature data

[0258] Output: Generated role-play scenario

[0259] Generate scenarios using neural networks or generative AI models (e.g., GPT-3, BERT). Specifically, use the TensorFlow library to create a neural network model with "tf.keras.models.Sequential()" to generate scenarios specifically for Salesperson A.

[0260] Step 5: Scenario Distribution

[0261] The server distributes the generated scenario to each salesperson's terminal.

[0262] Input: Generated role-play scenario

[0263] Output: Scenario sent to salesperson's terminal

[0264] The scenario is sent to the terminal using notification or file transfer. Specifically, the file is transferred using an HTTP request or FTP. For example, the server uses "requests.post(url, data=scenario data)" to send the scenario to Salesperson A's terminal.

[0265] Step 6: Role-playing and emotion recognition

[0266] The user performs role-playing based on the delivered scenario.

[0267] Input: Streamed role-play scenario

[0268] Output: Emotion data during role-playing

[0269] The device's built-in emotion engine (e.g., OpenFace or IBM Watson Tone Analyzer) analyzes the salesperson's facial expressions and tone of voice to recognize their emotions. For example, the device uses the OpenFace library to analyze emotions such as "stress" or "confidence" from the user's facial expression data.

[0270] Step 7: Scenario adjustment

[0271] The server dynamically adjusts the difficulty and content of the role-play scenario based on the emotional information obtained from the emotion engine.

[0272] Input: Emotion data

[0273] Output: A tailored role-play scenario

[0274] Use a reinforcement learning model to change the difficulty of a scenario in real time. For example, receive information that Salesperson A is feeling stressed and set the difficulty of the scenario to a lower level.

[0275] Step 8: Gather feedback

[0276] After the role-play is completed, the user inputs feedback into the terminal.

[0277] Input: User feedback

[0278] Output: Collected feedback data

[0279] The feedback includes the results of the role-play, impressions, and areas for improvement. Specifically, opinions are collected using Google Forms or an in-house developed feedback form. For example, users can enter their impressions, such as "My closing skills have improved," into the form displayed on their device.

[0280] Step 9: Feedback analysis and model updating

[0281] The server analyzes the collected feedback and updates the machine learning model.

[0282] Input: Collected feedback data

[0283] Output: An updated machine learning model

[0284] The feedback data is used to improve the accuracy of the model and reflected in the next scenario generation. Specifically, the feedback data is cleaned using the Pandas library and the model is retrained using Scikit-learn.

[0285] (Application example 2)

[0286] 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."

[0287] Conventional salesperson training systems lacked the ability to automatically generate scenarios optimized for each salesperson and dynamically adjust the difficulty and content based on their emotions. This made it difficult to provide effective training tailored to each salesperson's individual abilities and characteristics, limiting performance improvement. Furthermore, the inability to analyze salesperson emotions in real time and reflect them in the training content reduced the effectiveness of the training.

[0288] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for collecting sales data, market data, store data, and salesperson data; means for preprocessing the collected data; means for analyzing the preprocessed data and extracting features; means for generating role-play scenarios based on the extracted features; means for distributing the generated scenarios to a terminal including means for analyzing the emotions of the salesperson; means for the salesperson to perform role-plays based on the scenarios, adjusting the difficulty and content of the scenarios based on the emotion analysis results, and collecting feedback; and means for updating the machine learning model based on the collected feedback and generating the next scenario. This makes it possible to generate and adjust role-play scenarios optimized for individual salespersons, taking into account the emotions of the salespersons.

[0289] "Sales data" refers to data that indicates sales and transaction information of products related to sales activities.

[0290] "Market data" refers to data that reflects market conditions and trends, including the actions of competitors and consumer trends.

[0291] "Store data" is data related to a specific store, and includes store sales information, inventory status, customer demographics, and the like.

[0292] "Salesperson data" refers to data relating to individual salespersons, including sales performance, skill level, working hours, and the like.

[0293] "Preprocessing" refers to processes such as filling in missing values ​​in collected data, deleting duplicate data, and standardizing formats.

[0294] "Feature extraction" is the process of finding useful patterns and trends in the data being analyzed.

[0295] A "role play scenario" is a simulated sales situation scenario that a salesperson performs for learning and training purposes.

[0296] The "emotion engine" is an engine that analyzes the facial expressions and tone of voice of salespeople to recognize their emotions.

[0297] A "machine learning model" is an algorithm or system that learns from large amounts of data and performs pattern recognition and prediction.

[0298] "Feedback" refers to information such as evaluations, impressions, and areas for improvement collected from salespeople after the role-play is completed.

[0299] In this invention, a server, a terminal, and a user (salesperson) work together to realize a training system for salespersons. Specific embodiments for carrying out the invention are described below.

[0300] First, the server collects sales data, market data, store data, and salesperson data. This is done using APIs and database queries to obtain various data and store it in the database. It is also possible to obtain real-time data from external data sources.

[0301] Next, the server preprocesses the collected data, completing missing data, removing duplicate data, and standardizing the data format to prepare it for analysis. This preprocessing improves the accuracy of the analysis and enables more reliable feature extraction.

[0302] Based on the preprocessed data, the server performs data analysis and extracts characteristics of salespeople and stores. For example, it uses statistical methods and machine learning models to discover patterns such as which salespeople are effective on certain days or at certain times of the day, or which stores are effective in particular products.

[0303] After the features are extracted, the server generates role-play scenarios, leveraging neural networks and other generative AI models to generate scenarios optimized for each salesperson and store, including the skills and situations needed to improve the salesperson's performance.

[0304] The generated scenario is distributed from the server to the salesperson's device. For example, the scenario is sent to a device such as a smartphone or head-mounted display (HMD) via notification or file transfer. The salesperson then role-plays based on the distributed scenario.

[0305] At this time, the emotion engine installed in the device analyzes the salesperson's facial expressions and tone of voice to recognize their emotions. The emotion engine is responsible for analyzing in real time whether the salesperson is feeling stressed or relaxed. For example, if the server detects that the salesperson is feeling stressed, it can dynamically adjust the difficulty of the scenario based on that information.

[0306] After the role-play is completed, the user (salesperson) enters feedback into the terminal. The feedback includes the results of the role-play, impressions, and areas for improvement. This feedback is sent to the server and used to update the machine learning model. This makes it possible to provide even more accurate and effective scenarios the next time a scenario is generated.

[0307] For example, a scenario can be generated using the following prompt:

[0308] "Generate a scenario to improve closing skills for a female in her twenties who visits your store during lunchtime on a weekday. If the salesperson feels stressed, adjust the difficulty of the scenario to be less difficult."

[0309] Based on this prompt, the server can generate the optimal role-play scenario for specific conditions and skill sets.

[0310] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0311] Step 1: Data collection

[0312] The server collects sales data, market data, store data, and salesperson data using external APIs and database queries. It uses API endpoints and database connection information as input, retrieves the latest data from each data source, and stores it in the database. The output is a collection of the collected raw data.

[0313] Step 2: Data Preprocessing

[0314] The server preprocesses the collected raw data. It uses the collected raw data as input, complements missing data, removes duplicate data, and standardizes the format to make it suitable for analysis. The output is preprocessed clean data. Specifically, it finds defects in the data, complements them, and unifies inconsistent data formats.

[0315] Step 3: Data analysis

[0316] The server analyzes the preprocessed data and extracts features. It uses the preprocessed clean data as input. It uses statistical methods and machine learning models to analyze patterns of salespeople and stores. The output is feature data for each salesperson and store. Specifically, it finds patterns of salespeople who are strong on specific days of the week or at specific times of the day, or stores with high sales.

[0317] Step 4: Role-play scenario generation

[0318] The server generates role-play scenarios based on the extracted feature data. Utilizing neural networks and generative AI models, it designs scenarios optimized for each salesperson. Feature data and prompts are used as input. The output is the generated role-play scenario. Specifically, it automatically generates scenarios that strengthen the salesperson's weaknesses.

[0319] Step 5: Scenario distribution

[0320] The server distributes the generated scenario to the salesperson's terminal. It uses the generated role-play scenario as input. It sends the scenario to the terminal using notification or file transfer. The output is a notification or file that the scenario was distributed. Specifically, it sends a notification to the terminal so that the salesperson can check the received scenario.

[0321] Step 6: Sentiment Analysis

[0322] The device uses an emotion engine to analyze the salesperson's emotions during role-playing. Data on the salesperson's facial expressions and tone of voice are used as input. The emotion engine analyzes this data and recognizes emotional states such as stress or relaxation. The output is the recognized emotion data. Specifically, the device uses a camera and microphone to capture facial expressions and voice data in real time, which the emotion engine analyzes.

[0323] Step 7: Dynamically adjust the scenario

[0324] The server adjusts the difficulty and content of the scenario based on the emotional data sent from the terminal. Emotional data is used as input. If the salesperson is feeling stressed, the server adjusts the difficulty of the scenario, for example, by lowering it. The output is the adjusted scenario. Specifically, the server uses a difficulty adjustment algorithm to dynamically change the content of the scenario.

[0325] Step 8: Gather feedback

[0326] The terminal collects feedback from the salesperson after the role-play ends. The input is the feedback data entered by the user. The output is the collected feedback data. In concrete terms, the salesperson fills out a feedback form on the terminal, which is then sent to the server.

[0327] Step 9: Update the machine learning model

[0328] The server updates the machine learning model based on the collected feedback data. It uses the feedback data and past data as input. The updated model can provide more accurate scenarios the next time it generates a scenario. The output is an updated machine learning model. Specifically, it adds the new feedback data to the learning data and retrains the model.

[0329] 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.

[0330] 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.

[0331] 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.

[0332] [Second embodiment]

[0333] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0334] 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.

[0335] 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).

[0336] 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.

[0337] 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.

[0338] 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).

[0339] 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.

[0340] 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.

[0341] 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.

[0342] 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.

[0343] 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.

[0344] 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."

[0345] The system according to the present invention has the function of collecting sales data, market data, store data, and salesperson data, and then generating, distributing, and evaluating role-play scenarios based on this data. The specific processing of this program and a natural language explanation are given below.

[0346] First, the server collects data from external and internal sources, including store sales data, foot traffic, market data, and each salesperson's past performance data, and stores the collected data in a database.

[0347] The server then preprocesses the stored data, which includes cleansing the data, standardizing the format, and removing duplicate data. For example, it imputes missing data and removes irrelevant data. At this stage, the preprocessed data is prepared for analysis.

[0348] After the preprocessing is complete, the server performs data analysis. This analysis is carried out using statistical methods and machine learning algorithms to extract data characteristics. For example, analyzing the data of Salesperson A reveals that he tends to be strong at certain times of the day and weak at other times of the day.

[0349] Next, the server generates role-play scenarios optimized for each salesperson based on the extracted features. This scenario generation uses neural networks and other AI techniques. For example, a scenario is generated for Salesperson A to improve his or her closing skills with customers during weak times.

[0350] The generated scenario is distributed by the server to the terminal of the target salesperson. Once the scenario is distributed, the user (salesperson) uses the terminal to perform role-playing. After the role-playing is completed, the user enters feedback into the terminal. The feedback includes the effectiveness of the scenario, areas for improvement, and performance.

[0351] The collected feedback is sent to the server. The server analyzes the feedback data and updates the machine learning model. This update improves the accuracy of the next scenario generation. For example, by incorporating feedback from Salesperson A, a more effective next scenario is generated.

[0352] In this way, the system provides each salesperson with a customized role-play scenario, maximizing the effectiveness of the training. This system is expected to improve salesperson performance.

[0353] The processing flow will be explained below.

[0354] Specific processing flow of the program

[0355] Step 1:

[0356] The server collects data from external and internal sources, specifically sales data, market data, store data, and salesperson data, using APIs and database queries. For example, it retrieves sales data by store, market trend data, and salesperson historical performance data.

[0357] Step 2:

[0358] The server preprocesses the collected data. Specifically, it complements missing data, removes duplicate data, and standardizes the format. For example, it complements missing data with average values ​​and removes inappropriate data.

[0359] Step 3:

[0360] The server analyzes the preprocessed data. Specifically, it uses statistical techniques and machine learning models to extract patterns and features from the data. For example, it can discover whether there are strong or weak trends on certain days of the week or during certain times of the day from the data of Salesperson A.

[0361] Step 4:

[0362] The server generates role-play scenarios based on the extracted characteristics. Specifically, it uses neural networks and AI technology to generate scenarios tailored to the characteristics of each salesperson and store. For example, for Salesperson A, it creates a scenario that includes closing techniques that should be strengthened during times when Salesperson A is weak.

[0363] Step 5:

[0364] The server distributes the generated role-play scenario to the device of the salesperson. Specifically, the scenario is sent to the device using notifications or file transfers. For example, the scenario is pushed to the tablet of Salesperson A.

[0365] Step 6:

[0366] The user (salesperson) uses the terminal to carry out a role-play scenario. Specifically, the user reads the displayed scenario and acts based on it. For example, Salesperson A performs a role-play according to the scenario.

[0367] Step 7:

[0368] The user enters feedback on the role-play into the terminal. Specifically, the results and impressions are entered into a form on the terminal and sent. For example, Salesperson A enters feedback such as "The closing technique in the scenario was effective."

[0369] Step 8:

[0370] The server collects the feedback and stores it in a database. Specifically, it receives and stores the feedback data sent from the terminal. For example, the feedback of salesperson A is stored in the database.

[0371] Step 9:

[0372] The server analyzes the collected feedback and updates the machine learning model. Specifically, it learns from new data and improves the model's accuracy. For example, the model is retrained using feedback data from Salesperson A.

[0373] Through these steps, a series of processes are implemented to provide each salesperson with the most appropriate role-play scenario, thereby maximizing the effectiveness of sales training.

[0374] Example 1

[0375] 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."

[0376] Conventional salesperson training systems have the problem of being dependent on a uniform training program, making it difficult to customize based on the characteristics and performance of each individual salesperson. This means that it is difficult to effectively improve the performance of each salesperson, and the effectiveness of training is limited. In addition, there is a lack of a mechanism for flexibly improving the training program based on feedback, which creates challenges in long-term salesperson development.

[0377] 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.

[0378] In this invention, the server includes means for collecting sales data, market data, store data, and salesperson data, means for preprocessing the collected data, means for analyzing the preprocessed data using statistical techniques and machine learning algorithms to extract features, means for generating role-play scenarios using a neural network or the like based on the extracted features, means for distributing the generated scenarios to the terminals of each salesperson, means for the salespersons to use their terminals to perform role-plays and collect feedback, and means for analyzing the collected feedback and updating the machine learning model. This makes it possible to provide a training program customized for each salesperson and enable continuous improvement based on the feedback.

[0379] "Sales Data" refers to information relating to the sales of goods and services at a store.

[0380] "Market Data" means information about external market trends and economic factors.

[0381] "Store data" refers to information regarding the number of customers visiting the store and the store's operating status.

[0382] "Salesperson Data" means information about the accomplishments and performance of individual salespersons.

[0383] "Preprocessing" refers to processing the collected data such as filling in missing values, standardizing the format, and deleting inappropriate data.

[0384] "Analysis" refers to the extraction of features and patterns from data using statistical methods and machine learning algorithms.

[0385] A "role play scenario" is a training scenario used by salespeople for practice or simulation.

[0386] A "neural network" is a type of algorithm consisting of multiple layers of artificial neurons and used for pattern recognition and machine learning.

[0387] A "terminal" is a device, such as a computer or tablet, used by a salesperson.

[0388] "Feedback" refers to opinions and results regarding the effectiveness of the scenario and areas for improvement that the salesperson enters after performing the role-play scenario.

[0389] A "machine learning model" is an algorithm that analyzes data and makes predictions and classifications based on the results.

[0390] The system of the present invention optimizes salesperson training through multiple steps: data collection, preprocessing, analysis, scenario generation, scenario distribution, role-playing, feedback collection, feedback analysis, and updating of machine learning models. This system provides role-playing scenarios customized for each salesperson, enabling effective training.

[0391] First, the server collects sales data, market data, store data, and salesperson data. Specifically, sales data is obtained from the POS system, market data from an external API, store data from customer count sensors, and salesperson data from the HR system. These data are integrated and stored in a database on the server.

[0392] The server then pre-processes the collected data, which includes data cleansing (filling in missing data, removing irrelevant data), standardizing formats (for example, standardizing date formats), and removing duplicate data, making the data suitable for analysis.

[0393] Next, the server analyzes the preprocessed data. Statistical methods and machine learning algorithms are used for the analysis. Specific tools used include Python's Pandas and Scikit-learn. At this stage, data features and patterns are extracted. For example, analyzing the sales data of Salesperson A may reveal a strong sales trend during certain time periods.

[0394] After the features are extracted, the server uses a generative AI model such as a neural network to generate role-play scenarios based on these features. The scenarios include content that addresses specific weaknesses of the salesperson. For example, a scenario is generated for Salesperson A to improve his or her closing skills with customers during weak times.

[0395] The generated scenario is sent from the server to the salesperson's terminal. After the scenario is sent, the user (salesperson) uses the terminal to perform the role-play. Examples of specific prompts for the role-play include the following:

[0396] Prompt: "How can I approach a male customer in his 30s who visits my store between 3:00 PM and 5:00 PM on Sundays?"

[0397] After the user performs a role-play based on the scenario, they input feedback into the terminal, including the effectiveness of the scenario, areas for improvement, and performance.

[0398] The collected feedback is sent to a server, which analyzes it and updates the machine learning model based on the analysis results. This update improves the accuracy of the next scenario generation, making training for salespeople even more effective.

[0399] This system provides training scenarios optimized for each salesperson, maximizing the effectiveness of the training. For example, by implementing the scenarios provided by the system, Salesperson A's success rate in closing sales improved even during times when he was previously weak at dealing with customers.

[0400] In this way, the present invention improves the accuracy and effectiveness of salesperson training through a series of processes from data collection to model updating.

[0401] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0402] Step 1: Data collection

[0403] The server collects sales data, market data, store data, and salesperson data. Data collection is performed using API requests to obtain data from the POS system, external APIs, customer count sensors, and HR systems. Each piece of data is sent to the server in JSON or CSV format. The specific operation of data collection is for the server to periodically make API calls from each data source and store the obtained data in a database.

[0404] Input: Various external and internal data obtained by API requests

[0405] Output: Sales data, market data, store data, salesperson data stored in the database

[0406] Step 2: Data Preprocessing

[0407] The server preprocesses the collected data. Preprocessing includes data cleansing, format standardization, and removal of duplicate data. Missing data is filled in with the average value from the past, and inappropriate data is removed. To standardize the format, the date format is standardized to "YYYY-MM-DD." Specifically, the server uses the Python Pandas library to manipulate data frames and perform cleansing and format standardization.

[0408] Input: Raw data stored in the database (sales data, market data, store data, salesperson data)

[0409] Output: Preprocessed data (cleaned data)

[0410] Step 3: Data analysis

[0411] The server analyzes the preprocessed data. Statistical methods and machine learning algorithms are used for this analysis. Specific methods include regression analysis and clustering using Python's Scikit-learn. The server extracts each salesperson's characteristics and trends based on the salesperson data. For example, it can discover the sales patterns of salesperson A, who is strong during specific time periods.

[0412] Input: Preprocessed data (cleaned data)

[0413] Output: Analyzed feature data (statistical results and machine learning model output)

[0414] Step 4: Scenario generation

[0415] The server generates role-play scenarios based on the analyzed feature data. Using neural networks and generative AI models, optimal scenarios are created for each salesperson. For example, a scenario is generated for Salesperson A to improve his closing skills with customers during a specific time period. The generated scenarios are converted into natural-sounding sentences using natural language generation (NLG) technology.

[0416] Input: Analyzed feature data (features for each salesperson)

[0417] Output: Generated role-play scenario (text format)

[0418] Step 5: Scenario distribution and execution

[0419] The server distributes the generated scenario to each salesperson's device. The salesperson (user) uses the device to execute the scenario. For example, Salesperson A performs a customer service simulation while reading the prompts displayed on the device. Push notifications and APIs are used to distribute the scenario.

[0420] Input: Generated role-play scenario (text format)

[0421] Output: Scenario delivered to the terminal (text displayed on the salesperson's terminal)

[0422] Step 6: Gather feedback

[0423] After the user (salesperson) performs the role-play scenario, they input their feedback into the terminal. The feedback includes the effectiveness of the scenario, areas for improvement, and performance. The terminal then sends this feedback data to the server.

[0424] Input: Feedback entered by the salesperson into the terminal (text format)

[0425] Output: Feedback data sent to the server (JSON or CSV format)

[0426] Step 7: Analyze feedback and update the model

[0427] The server analyzes the collected feedback data and updates the machine learning model. For analysis, the server uses a Python library to retrain the model parameters based on the feedback data, which makes the next scenario generation more accurate.

[0428] Input: Feedback data stored on the server (text, JSON, CSV)

[0429] Output: Updated machine learning model (improved AI model)

[0430] This series of processes systematically supports salespeople in improving their performance, and continuous learning effects can be expected.

[0431] (Application example 1)

[0432] 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."

[0433] Aiming to improve the performance of salespeople in brick-and-mortar stores, there is a need to effectively improve sales skills, especially closing skills, by providing training scenarios tailored to each individual salesperson. Traditional training methods have struggled to effectively incorporate feedback and ensure continuous improvement. Furthermore, there are limited means of evaluating training content in real time using technologies such as voice recognition. Therefore, a system that can accurately address each salesperson's weaknesses is needed.

[0434] 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.

[0435] In this invention, the server includes means for collecting sales data, market data, store data, and salesperson data, means for preprocessing the collected data, means for analyzing the preprocessed data and extracting features, means for generating role-play scenarios based on the extracted features, means for distributing the generated scenarios to terminals of each salesperson, means for the salesperson to perform role-plays based on the scenarios and collect feedback, means for updating a machine learning model based on the collected feedback, means for generating scenarios for the salesperson to improve their closing skills for a specific product, means for the terminal to automatically evaluate responses in the role-plays using a voice recognition function, and means for using the feedback data to reflect the feedback data in the generation of the next scenario. This makes it possible to generate and evaluate training scenarios tailored to individual salespersons and reflect the feedback.

[0436] "Sales data" is information relating to product sales, including sales quantity, sales amount, sales date, sales time, and the like.

[0437] "Market data" is information about the current market situation, such as competitor trends, fluctuations in product prices, and consumer purchasing intentions.

[0438] "Store data" is information about a specific store, including the number of customers, inventory status, store layout, business hours, and the like.

[0439] "Salesperson data" is information about each salesperson, including past sales performance, customer service skills, working hours, and training history.

[0440] "Preprocessing" refers to preparing the collected raw data in a format that can be analyzed, and includes steps such as data cleansing, format standardization, and missing value imputation.

[0441] "Feature extraction" refers to finding important patterns and trends through data analysis, using statistical analysis and machine learning algorithms to reveal the essential parts of the data.

[0442] A "role-play scenario" is a training scenario that allows salespeople to improve their skills by interacting with virtual customers, and includes specific customer service situations and troubleshooting methods.

[0443] "Terminal" refers to the device used by the salesperson, which may be a smartphone, tablet, or other portable computing device.

[0444] "Feedback" refers to salespeople providing their opinions on the effectiveness of the scenario and areas for improvement after the role-play, and this information is collected and used as valuable data to improve the system.

[0445] A "machine learning model" is an algorithm that generates patterns and predictions based on past data. It is used to update the model based on collected feedback and improve the quality of training scenarios.

[0446] The "voice recognition function" is a technology that allows the terminal to analyze the salesperson's voice in real time and process the spoken content as text data, and is used to automatically evaluate the responses in role-playing.

[0447] "Closing skills" are techniques for getting customers to make a purchase decision at the final stage of sales negotiations or customer service, and are important abilities that include persuasiveness and communication skills.

[0448] The system according to the present invention has the function of collecting sales data, market data, store data, and salesperson data, and then generating, distributing, and evaluating role-play scenarios based on these data. The specific processing of this system is explained below in natural language.

[0449] First, the server collects data from external and internal sources, including store sales data, foot traffic, market data, and each salesperson's past performance data, and stores the collected data in a database.

[0450] The server then preprocesses the stored data, which includes cleansing the data, standardizing the format, and removing duplicate data. For example, it imputes missing data and removes irrelevant data. At this stage, the preprocessed data is prepared for analysis.

[0451] After the preprocessing is complete, the server performs data analysis. This analysis is carried out using statistical methods and machine learning algorithms to extract data characteristics. For example, analyzing the data of Salesperson A reveals that he tends to be strong at certain times of the day and weak at other times of the day.

[0452] Next, the server generates role-play scenarios optimized for each salesperson based on the extracted features. Neural networks and other AI techniques are used to generate these scenarios. For example, a scenario is generated for Salesperson A to improve his closing skills with customers during weak times. The prompt might be something like, "Salesperson A is having trouble closing sales for a specific product. Please generate a role-play scenario to improve his skills in promoting this product."

[0453] The generated scenario is distributed by the server to the target salesperson's terminal. Once the scenario is distributed, the salesperson uses the terminal to perform role-playing. For example, Salesperson A receives a scenario to practice closing a specific product, and then interacts with a virtual customer based on that scenario.

[0454] After the role-play, the salesperson enters feedback into the terminal, including the effectiveness of the scenario, areas for improvement, and performance. The terminal may also automatically evaluate the responses during the role-play using voice recognition.

[0455] The collected feedback is sent to the server. The server analyzes the feedback data and updates the machine learning model. This update improves the accuracy of the next scenario generation. For example, by incorporating feedback from Salesperson A, a more effective next scenario is generated.

[0456] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0457] Step 1:

[0458] First, the server collects data from external and internal sources, including store sales data, customer traffic, market data, and each salesperson's past performance data. The input is data from the sales system, customer management system, and market information system, and the output is stored in a database.

[0459] Step 2:

[0460] Next, the server preprocesses the stored data. This includes data cleansing, formatting standardization, and deleting duplicate data. The input is the raw data collected in step 1, and the output is data in a clean, analyzable format. Specifically, missing values ​​are imputed, inappropriate data is deleted, and the format is standardized.

[0461] Step 3:

[0462] The server then performs data analysis based on the preprocessed data. Statistical methods and machine learning algorithms are used for the analysis. For example, based on the data of Salesperson A, strengths and weaknesses in specific time periods can be extracted. The input is the preprocessed data, and the output is an index showing the characteristics of the data.

[0463] Step 4:

[0464] Next, the server generates role-play scenarios optimized for each salesperson based on the extracted features. Neural networks and other AI techniques are used to generate the scenarios. For example, a prompt such as "Salesperson A is struggling to close a specific product. Please generate a role-play scenario to help him improve his skills in promoting this product" can be input, generating a scenario focused on specific closing skills. The output is a specific role-play scenario.

[0465] Step 5:

[0466] The generated scenario is delivered by the server to the target salesperson's terminal. The input is the generated role-play scenario, and the output is the delivered scenario. The salesperson can receive the role-play scenario on their terminal.

[0467] Step 6:

[0468] After the scenario is delivered, the salesperson uses the terminal to perform the role-play. Following the scenario on the terminal, the salesperson interacts with a virtual customer. The input is the delivered scenario, and the output is the result of the role-play.

[0469] Step 7:

[0470] After the role-play, the salesperson enters their feedback into the terminal. The feedback includes the effectiveness of the scenario, areas for improvement, and performance. The input is the results of the role-play and the salesperson's opinions, and the output is structured feedback data.

[0471] Step 8:

[0472] The collected feedback is sent to the server, which analyzes the feedback data and updates the machine learning model. The input is the feedback data, and the output is an updated machine learning model. This improves the accuracy of the next scenario generation.

[0473] Through these steps, a series of processes are carried out, providing each salesperson with a role-play scenario that is optimized for them.

[0474] 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.

[0475] This invention relates to a system that collects and analyzes sales data, market data, store data, and salesperson data, and generates role-play scenarios based on these data. Furthermore, by combining this with an emotion engine that recognizes the emotions of users (salespersons), more effective training can be achieved.

[0476] First, the server collects sales data, market data, store data, and salesperson data from external and internal sources, using APIs and database queries to retrieve various data and store it in a database.

[0477] Next, the server preprocesses the collected data. Since the collected data is often incomplete, missing data is filled in, duplicate data is removed, and the format is standardized. This prepares the data for analysis.

[0478] The server then performs data analysis based on the preprocessed data. Statistical methods and machine learning models are used to extract the characteristics of each salesperson and store. For example, patterns can be discovered, such as strengths on certain days of the week or at certain times of the day, but weaknesses on other days or at certain times of the day.

[0479] The server then generates role-play scenarios based on the extracted features. Using neural networks and other AI techniques, scenarios optimized for each salesperson and store are generated. For example, for Salesperson A, a scenario is created to strengthen closing skills to overcome weak performance.

[0480] The generated scenario is delivered from the server to the target salesperson's terminal. The scenario is sent to the terminal using notifications or file transfers. The user (salesperson) then performs a role-play based on the delivered scenario.

[0481] When the user is performing a role-play, the emotion engine installed in the device analyzes the salesperson's facial expressions and tone of voice to recognize their emotions. For example, if Salesperson A is feeling stressed, this information will be detected by the emotion engine.

[0482] The emotional information analyzed by the emotion engine is sent to the server. The server then dynamically adjusts the difficulty and content of the role-play scenario based on this information. For example, if Salesperson A feels stressed by a difficult scenario, the difficulty of the scenario can be lowered.

[0483] After the role-playing is completed, the user inputs feedback into the terminal. The feedback includes the results of the role-playing, impressions, and areas for improvement. The server collects this feedback and stores it in a database.

[0484] The collected feedback is then analyzed by the server again to update the machine learning model, which will enable it to provide more accurate and effective scenarios the next time it generates them.

[0485] In this way, by providing customized role-play scenarios that take into account the emotions of salespeople, it is expected that the effectiveness of training will be maximized and salespeople's performance will improve.

[0486] The processing flow will be explained below.

[0487] MODE FOR CARRYING OUT THE INVENTION

[0488] Step 1:

[0489] The server collects sales data, market data, store data, and salesperson data from external and internal sources. Specifically, it retrieves various data using APIs and database queries and stores it in a database. For example, store sales data and market trend data can be retrieved through an API and stored in the appropriate database tables.

[0490] Step 2:

[0491] The server preprocesses the collected data. Specific operations include filling in missing data, deleting duplicate data, and standardizing formats. For example, if there are gaps in sales data, it fills in the gaps with the average value from the past, and if the data format is inconsistent, it converts it into a unified format.

[0492] Step 3:

[0493] The server then performs data analysis using the preprocessed data. Specifically, statistical methods and machine learning models are used to extract patterns and features from the data. For example, analysis can be performed to discover a trend in which Salesperson A's performance is weak during certain time periods.

[0494] Step 4:

[0495] The server generates role-play scenarios based on the extracted features. As a specific example, neural networks and AI technology are used to create scenarios that strengthen salespeople's weaknesses. For example, a role-play scenario is generated to strengthen Salesperson A's closing skills when dealing with customers during weekdays.

[0496] Step 5:

[0497] The server generates a role-play scenario and distributes it to the salesperson's device. Specifically, it uses notification and file transfer functions to send the scenario to the salesperson's tablet or smartphone.

[0498] Step 6:

[0499] The user (salesperson) uses the terminal to perform role-playing. Specifically, the user follows a scenario displayed on the terminal to simulate customer interactions. For example, a role-playing session is conducted in which Salesperson A tries out a new closing technique.

[0500] Step 7:

[0501] During the role-play, the device's built-in emotion engine recognizes the user's emotions. For example, it uses a camera and microphone to analyze the salesperson's facial expressions and tone of voice, collecting emotional data such as stress and satisfaction.

[0502] Step 8:

[0503] The server receives the emotion data sent from the emotion engine and dynamically adjusts the difficulty and content of the role-play scenario. For example, if Salesperson A is feeling stressed, the server adjusts the difficulty of the next scenario based on this information.

[0504] Step 9:

[0505] The user enters feedback on the role-play into the terminal. Specifically, after the role-play is completed, the user enters the results, impressions, and points for improvement into the feedback form and submits it. For example, Salesperson A enters feedback such as, "The closing techniques in the scenario were effective."

[0506] Step 10:

[0507] The server analyzes the collected feedback and stores it in a database. Specifically, it receives and analyzes the feedback data to update the machine learning model. This data is reflected in the next scenario generation. For example, the feedback from Salesperson A is used to learn how to create a more precise scenario for the next time.

[0508] This process maximizes the effectiveness of training by providing each salesperson with tailored role-play scenarios.

[0509] Example 2

[0510] 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."

[0511] Conventional salesperson training systems have difficulty providing real-time feedback on salespersons' actual performance and emotions, making it impossible to provide training scenarios optimized for each individual salesperson. Furthermore, there was no system in place to properly collect and analyze training feedback and reflect it in the next training session. As a result, the effectiveness of training was limited.

[0512] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0513] In this invention, the server includes means for collecting sales data, market data, store data, and salesperson data, means for preprocessing the collected data, means for analyzing the preprocessed data and extracting features, means for generating role-play scenarios based on the extracted features using a neural network or other generative AI model, means for distributing the generated scenarios to each salesperson's terminal, means for the salesperson to perform role-plays based on the scenarios and for recognizing emotions using an emotion engine of the terminal, means for dynamically adjusting the difficulty and content of the role-play scenarios based on the emotion information, means for collecting feedback after the scenarios are performed, and means for updating the machine learning model based on the collected feedback. This makes it possible to provide effective training scenarios based on the salesperson's actual performance and emotions and to reflect the feedback in real time.

[0514] "Sales Data" refers to information related to the sale of goods or services, including sales quantity, sales amount, sales date and time, customer information, etc.

[0515] "Market data" refers to information related to market conditions, including the balance of supply and demand, price fluctuations, competitive conditions, and economic indicators.

[0516] "Store data" refers to information related to a specific store, including the store's location, size, business hours, staff information, sales performance, etc.

[0517] "Salesperson Data" means information related to individual salespersons, including their skills, experience, ratings, and performance.

[0518] "Preprocessing" refers to tasks such as data cleaning and format conversion to convert collected data into an analyzable format.

[0519] "Feature extraction" refers to the process of finding important patterns and relationships in pre-processed data.

[0520] "Generative AI models" refer to technologies that use neural networks and other artificial intelligence algorithms to generate new information and scenarios from data.

[0521] A "role-play scenario" refers to a scenario created for the purpose of training salespeople to practice sales activities in specific situations.

[0522] "Terminal" refers to the digital device used by the salesperson, including PCs, tablets, smartphones, etc.

[0523] An "emotion engine" refers to software or hardware that analyzes and recognizes emotions from a salesperson's facial expressions, tone of voice, etc.

[0524] "Feedback" refers to comments and evaluations provided by salespeople after the role-play, including information on the effectiveness of the training and areas for improvement.

[0525] A "machine learning model" is an algorithm that makes predictions or classifications based on data and is trained using training data.

[0526] The present invention relates to a system that collects and analyzes sales data, market data, store data, and salesperson data, and generates role-play scenarios based on these data. Furthermore, by combining this system with an emotion engine that recognizes the emotions of users (salespersons), it is possible to achieve more effective training. Specific embodiments are described below.

[0527] First, the server collects sales data, market data, store data, and salesperson data from external and internal sources. To do this, it uses APIs and database queries to retrieve various data and stores it in a database. Specifically, it uses RESTful APIs and SQL queries. For example, the server retrieves market data from "https: / / api.example.com / marketdata" and retrieves sales data with the SQL query "SELECT FROM sales WHERE date='2023-10-01'".

[0528] Next, the data collected by the server is preprocessed. As the data is often incomplete as it is, missing data is complemented, duplicate data is deleted, and the format is standardized. Specifically, using the Python Pandas library, duplicate data is deleted using the ".drop_duplicates()" method, and missing data is complemented using the ".fillna(method='ffill')" method.

[0529] The server performs data analysis based on the preprocessed data. It uses statistical methods and machine learning models (e.g., regression analysis, clustering) to extract the characteristics of each salesperson and store. For example, the server uses the Scikit-learn library to analyze the relationship between sales data and days of the week using the "LinearRegression()" model.

[0530] Next, the server generates role-play scenarios based on the analysis results. It uses neural networks or generative AI models (e.g., GPT-3, BERT) to generate scenarios optimized for each salesperson and store. Specifically, it uses the TensorFlow library to create a neural network model with "tf.keras.models.Sequential()" to generate a scenario specifically for Salesperson A.

[0531] The generated scenario is distributed from the server to the terminal of the target salesperson. The scenario is sent to the terminal using notification or file transfer. Specifically, the file is transferred using an HTTP request or FTP. For example, the server uses "requests.post(url, data=scenario data)" to send the scenario to Salesperson A's terminal.

[0532] The user (salesperson) performs a role-play based on the delivered scenario. During the role-play, the device's built-in emotion engine (e.g., OpenFace or IBM Watson Tone Analyzer) analyzes the salesperson's facial expressions and tone of voice to recognize their emotions. For example, the device uses the OpenFace library to analyze emotions such as "stress" or "confidence" from the user's facial expression data.

[0533] The emotional information analyzed by the emotion engine is sent to the server. The server uses this information to dynamically adjust the difficulty and content of the role-play scenario. Specifically, it uses a reinforcement learning model to change the difficulty of the scenario in real time. For example, if it receives information that Salesperson A is feeling stressed, it will set the difficulty of the scenario to a lower level.

[0534] After the role-play is completed, the user enters feedback into the device. The feedback includes the results of the role-play, their impressions, and areas for improvement. Specifically, opinions are collected using Google Forms or a feedback form developed in-house. For example, the user may enter their impression that "my closing skills have improved" into the form displayed on the device.

[0535] The server analyzes the collected feedback and updates the machine learning model. The feedback data is used to improve the model's accuracy and is reflected in the next scenario generation. Specifically, the feedback data is cleaned using the Pandas library and the model is retrained using Scikit-learn.

[0536] Prompt Sentence Examples

[0537] "You are the server of a system that collects and analyzes sales data, market data, store data, and salesperson data, and generates role-play scenarios based on this data. Specifically, please perform the following tasks:

[0538] 1. Collect data using API queries or SQL queries and store it in a database.

[0539] 2. Preprocess the data using the Pandas library.

[0540] 3. Analyze the data using a machine learning model to extract the characteristics of each salesperson.

[0541] 4. Generate role-play scenarios using TensorFlow.

[0542] 5. The scenario is delivered to the user's device via an HTTP request.

[0543] 6. The user performs a role-play based on the scenario, and the device's emotion engine recognizes emotions.

[0544] 7. Adjust the difficulty of the scenario based on emotional data.

[0545] 8. Collect user feedback and incorporate it into the next scenario generation.

[0546] For example, if Salesperson A is feeling stressed, adjust the scenario to be less difficult and use the feedback data to update the machine learning model for effective training.

[0547] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0548] Step 1: Data collection

[0549] The server obtains sales data, market data, store data, and salesperson data from external and internal sources.

[0550] Input: API query, SQL query

[0551] Output: Collected data (sales data, market data, store data, salesperson data)

[0552] The server retrieves market information using a RESTful API and pulls store information from an internal database using SQL queries. For example, it retrieves market data from "https: / / api.example.com / marketdata" and retrieves sales data with the SQL query "SELECT FROM sales WHERE date='2023-10-01'".

[0553] Step 2: Data Preprocessing

[0554] The server preprocesses the collected data.

[0555] Input: Collected data

[0556] Output: Preprocessed data

[0557] The server uses the Python Pandas library to clean the data. Specifically, it removes duplicate data using the ".drop_duplicates()" method, fills missing data using the ".fillna(method='ffill')" method, and standardizes the data format.

[0558] Step 3: Data analysis

[0559] The server performs analysis based on the preprocessed data and extracts features.

[0560] Input: Preprocessed data

[0561] Output: Extracted feature data

[0562] The server uses the Scikit-learn library to analyze the data using statistical methods and machine learning models (e.g., regression analysis, clustering). Specifically, it uses the "LinearRegression()" model to analyze the relationship between sales data and days of the week and extract features.

[0563] Step 4: Role-play scenario generation

[0564] The server generates a role-play scenario based on the extracted features.

[0565] Input: Extracted feature data

[0566] Output: Generated role-play scenario

[0567] Generate scenarios using neural networks or generative AI models (e.g., GPT-3, BERT). Specifically, use the TensorFlow library to create a neural network model with "tf.keras.models.Sequential()" to generate scenarios specifically for Salesperson A.

[0568] Step 5: Scenario Distribution

[0569] The server distributes the generated scenario to each salesperson's terminal.

[0570] Input: Generated role-play scenario

[0571] Output: Scenario sent to salesperson's terminal

[0572] The scenario is sent to the terminal using notification or file transfer. Specifically, the file is transferred using an HTTP request or FTP. For example, the server uses "requests.post(url, data=scenario data)" to send the scenario to Salesperson A's terminal.

[0573] Step 6: Role-playing and emotion recognition

[0574] The user performs role-playing based on the delivered scenario.

[0575] Input: Streamed role-play scenario

[0576] Output: Emotion data during role-playing

[0577] The device's built-in emotion engine (e.g., OpenFace or IBM Watson Tone Analyzer) analyzes the salesperson's facial expressions and tone of voice to recognize their emotions. For example, the device uses the OpenFace library to analyze emotions such as "stress" or "confidence" from the user's facial expression data.

[0578] Step 7: Scenario adjustment

[0579] The server dynamically adjusts the difficulty and content of the role-play scenario based on the emotional information obtained from the emotion engine.

[0580] Input: Emotion data

[0581] Output: A tailored role-play scenario

[0582] Use a reinforcement learning model to change the difficulty of a scenario in real time. For example, receive information that Salesperson A is feeling stressed and set the difficulty of the scenario to a lower level.

[0583] Step 8: Gather feedback

[0584] After the role-play is completed, the user inputs feedback into the terminal.

[0585] Input: User feedback

[0586] Output: Collected feedback data

[0587] The feedback includes the results of the role-play, impressions, and areas for improvement. Specifically, opinions are collected using Google Forms or an in-house developed feedback form. For example, users can enter their impressions, such as "My closing skills have improved," into the form displayed on their device.

[0588] Step 9: Feedback analysis and model updating

[0589] The server analyzes the collected feedback and updates the machine learning model.

[0590] Input: Collected feedback data

[0591] Output: An updated machine learning model

[0592] The feedback data is used to improve the accuracy of the model and reflected in the next scenario generation. Specifically, the feedback data is cleaned using the Pandas library and the model is retrained using Scikit-learn.

[0593] (Application example 2)

[0594] 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."

[0595] Conventional salesperson training systems lacked the ability to automatically generate scenarios optimized for each salesperson and dynamically adjust the difficulty and content based on their emotions. This made it difficult to provide effective training tailored to each salesperson's individual abilities and characteristics, limiting performance improvement. Furthermore, the inability to analyze salesperson emotions in real time and reflect them in the training content reduced the effectiveness of the training.

[0596] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for collecting sales data, market data, store data, and salesperson data; means for preprocessing the collected data; means for analyzing the preprocessed data and extracting features; means for generating role-play scenarios based on the extracted features; means for distributing the generated scenarios to a terminal including means for analyzing the emotions of the salesperson; means for the salesperson to perform role-plays based on the scenarios, adjusting the difficulty and content of the scenarios based on the emotion analysis results, and collecting feedback; and means for updating the machine learning model based on the collected feedback and generating the next scenario. This makes it possible to generate and adjust role-play scenarios optimized for individual salespersons, taking into account the emotions of the salespersons.

[0597] "Sales data" refers to data that indicates sales and transaction information of products related to sales activities.

[0598] "Market data" refers to data that reflects market conditions and trends, including the actions of competitors and consumer trends.

[0599] "Store data" is data related to a specific store, and includes store sales information, inventory status, customer demographics, and the like.

[0600] "Salesperson data" refers to data relating to individual salespersons, including sales performance, skill level, working hours, and the like.

[0601] "Preprocessing" refers to processes such as filling in missing values ​​in collected data, deleting duplicate data, and standardizing formats.

[0602] "Feature extraction" is the process of finding useful patterns and trends in the data being analyzed.

[0603] A "role play scenario" is a simulated sales situation scenario that a salesperson performs for learning and training purposes.

[0604] The "emotion engine" is an engine that analyzes the facial expressions and tone of voice of salespeople to recognize their emotions.

[0605] A "machine learning model" is an algorithm or system that learns from large amounts of data and performs pattern recognition and prediction.

[0606] "Feedback" refers to information such as evaluations, impressions, and areas for improvement collected from salespeople after the role-play is completed.

[0607] In this invention, a server, a terminal, and a user (salesperson) work together to realize a training system for salespersons. Specific embodiments for carrying out the invention are described below.

[0608] First, the server collects sales data, market data, store data, and salesperson data. This is done using APIs and database queries to obtain various data and store it in the database. It is also possible to obtain real-time data from external data sources.

[0609] Next, the server preprocesses the collected data, completing missing data, removing duplicate data, and standardizing the data format to prepare it for analysis. This preprocessing improves the accuracy of the analysis and enables more reliable feature extraction.

[0610] Based on the preprocessed data, the server performs data analysis and extracts characteristics of salespeople and stores. For example, it uses statistical methods and machine learning models to discover patterns such as which salespeople are effective on certain days or at certain times of the day, or which stores are effective in particular products.

[0611] After the features are extracted, the server generates role-play scenarios, leveraging neural networks and other generative AI models to generate scenarios optimized for each salesperson and store, including the skills and situations needed to improve the salesperson's performance.

[0612] The generated scenario is distributed from the server to the salesperson's device. For example, the scenario is sent to a device such as a smartphone or head-mounted display (HMD) via notification or file transfer. The salesperson then role-plays based on the distributed scenario.

[0613] At this time, the emotion engine installed in the device analyzes the salesperson's facial expressions and tone of voice to recognize their emotions. The emotion engine is responsible for analyzing in real time whether the salesperson is feeling stressed or relaxed. For example, if the server detects that the salesperson is feeling stressed, it can dynamically adjust the difficulty of the scenario based on that information.

[0614] After the role-play is completed, the user (salesperson) enters feedback into the terminal. The feedback includes the results of the role-play, impressions, and areas for improvement. This feedback is sent to the server and used to update the machine learning model. This makes it possible to provide even more accurate and effective scenarios the next time a scenario is generated.

[0615] For example, a scenario can be generated using the following prompt:

[0616] "Generate a scenario to improve closing skills for a female in her twenties who visits your store during lunchtime on a weekday. If the salesperson feels stressed, adjust the difficulty of the scenario to be less difficult."

[0617] Based on this prompt, the server can generate the optimal role-play scenario for specific conditions and skill sets.

[0618] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0619] Step 1: Data collection

[0620] The server collects sales data, market data, store data, and salesperson data using external APIs and database queries. It uses API endpoints and database connection information as input, retrieves the latest data from each data source, and stores it in the database. The output is a collection of the collected raw data.

[0621] Step 2: Data Preprocessing

[0622] The server preprocesses the collected raw data. It uses the collected raw data as input, complements missing data, removes duplicate data, and standardizes the format to make it suitable for analysis. The output is preprocessed clean data. Specifically, it finds defects in the data, complements them, and unifies inconsistent data formats.

[0623] Step 3: Data analysis

[0624] The server analyzes the preprocessed data and extracts features. It uses the preprocessed clean data as input. It uses statistical methods and machine learning models to analyze patterns of salespeople and stores. The output is feature data for each salesperson and store. Specifically, it finds patterns of salespeople who are strong on specific days of the week or at specific times of the day, or stores with high sales.

[0625] Step 4: Role-play scenario generation

[0626] The server generates role-play scenarios based on the extracted feature data. Utilizing neural networks and generative AI models, it designs scenarios optimized for each salesperson. Feature data and prompts are used as input. The output is the generated role-play scenario. Specifically, it automatically generates scenarios that strengthen the salesperson's weaknesses.

[0627] Step 5: Scenario distribution

[0628] The server distributes the generated scenario to the salesperson's terminal. It uses the generated role-play scenario as input. It sends the scenario to the terminal using notification or file transfer. The output is a notification or file that the scenario was distributed. Specifically, it sends a notification to the terminal so that the salesperson can check the received scenario.

[0629] Step 6: Sentiment Analysis

[0630] The device uses an emotion engine to analyze the salesperson's emotions during role-playing. Data on the salesperson's facial expressions and tone of voice are used as input. The emotion engine analyzes this data and recognizes emotional states such as stress or relaxation. The output is the recognized emotion data. Specifically, the device uses a camera and microphone to capture facial expressions and voice data in real time, which the emotion engine analyzes.

[0631] Step 7: Dynamically adjust the scenario

[0632] The server adjusts the difficulty and content of the scenario based on the emotional data sent from the terminal. Emotional data is used as input. If the salesperson is feeling stressed, the server adjusts the difficulty of the scenario, for example, by lowering it. The output is the adjusted scenario. Specifically, the server uses a difficulty adjustment algorithm to dynamically change the content of the scenario.

[0633] Step 8: Gather feedback

[0634] The terminal collects feedback from the salesperson after the role-play ends. The input is the feedback data entered by the user. The output is the collected feedback data. In concrete terms, the salesperson fills out a feedback form on the terminal, which is then sent to the server.

[0635] Step 9: Update the machine learning model

[0636] The server updates the machine learning model based on the collected feedback data. It uses the feedback data and past data as input. The updated model can provide more accurate scenarios the next time it generates a scenario. The output is an updated machine learning model. Specifically, it adds the new feedback data to the learning data and retrains the model.

[0637] 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.

[0638] 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.

[0639] 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.

[0640] [Third embodiment]

[0641] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0642] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0643] 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).

[0644] 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.

[0645] 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.

[0646] 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).

[0647] 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.

[0648] 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.

[0649] 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.

[0650] 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.

[0651] 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.

[0652] 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."

[0653] The system according to the present invention has the function of collecting sales data, market data, store data, and salesperson data, and then generating, distributing, and evaluating role-play scenarios based on this data. The specific processing of this program and a natural language explanation are given below.

[0654] First, the server collects data from external and internal sources, including store sales data, foot traffic, market data, and each salesperson's past performance data, and stores the collected data in a database.

[0655] The server then preprocesses the stored data, which includes cleansing the data, standardizing the format, and removing duplicate data. For example, it imputes missing data and removes irrelevant data. At this stage, the preprocessed data is prepared for analysis.

[0656] After the preprocessing is complete, the server performs data analysis. This analysis is carried out using statistical methods and machine learning algorithms to extract data characteristics. For example, analyzing the data of Salesperson A reveals that he tends to be strong at certain times of the day and weak at other times of the day.

[0657] Next, the server generates role-play scenarios optimized for each salesperson based on the extracted features. This scenario generation uses neural networks and other AI techniques. For example, a scenario is generated for Salesperson A to improve his or her closing skills with customers during weak times.

[0658] The generated scenario is distributed by the server to the terminal of the target salesperson. Once the scenario is distributed, the user (salesperson) uses the terminal to perform role-playing. After the role-playing is completed, the user enters feedback into the terminal. The feedback includes the effectiveness of the scenario, areas for improvement, and performance.

[0659] The collected feedback is sent to the server. The server analyzes the feedback data and updates the machine learning model. This update improves the accuracy of the next scenario generation. For example, by incorporating feedback from Salesperson A, a more effective next scenario is generated.

[0660] In this way, the system provides each salesperson with a customized role-play scenario, maximizing the effectiveness of the training. This system is expected to improve salesperson performance.

[0661] The processing flow will be explained below.

[0662] Specific processing flow of the program

[0663] Step 1:

[0664] The server collects data from external and internal sources, specifically sales data, market data, store data, and salesperson data, using APIs and database queries. For example, it retrieves sales data by store, market trend data, and salesperson historical performance data.

[0665] Step 2:

[0666] The server preprocesses the collected data. Specifically, it complements missing data, removes duplicate data, and standardizes the format. For example, it complements missing data with average values ​​and removes inappropriate data.

[0667] Step 3:

[0668] The server analyzes the preprocessed data. Specifically, it uses statistical techniques and machine learning models to extract patterns and features from the data. For example, it can discover whether there are strong or weak trends on certain days of the week or during certain times of the day from the data of Salesperson A.

[0669] Step 4:

[0670] The server generates role-play scenarios based on the extracted characteristics. Specifically, it uses neural networks and AI technology to generate scenarios tailored to the characteristics of each salesperson and store. For example, for Salesperson A, it creates a scenario that includes closing techniques that should be strengthened during times when Salesperson A is weak.

[0671] Step 5:

[0672] The server distributes the generated role-play scenario to the device of the salesperson. Specifically, the scenario is sent to the device using notifications or file transfers. For example, the scenario is pushed to the tablet of Salesperson A.

[0673] Step 6:

[0674] The user (salesperson) uses the terminal to carry out a role-play scenario. Specifically, the user reads the displayed scenario and acts based on it. For example, Salesperson A performs a role-play according to the scenario.

[0675] Step 7:

[0676] The user enters feedback on the role-play into the terminal. Specifically, the results and impressions are entered into a form on the terminal and sent. For example, Salesperson A enters feedback such as "The closing technique in the scenario was effective."

[0677] Step 8:

[0678] The server collects the feedback and stores it in a database. Specifically, it receives and stores the feedback data sent from the terminal. For example, the feedback of salesperson A is stored in the database.

[0679] Step 9:

[0680] The server analyzes the collected feedback and updates the machine learning model. Specifically, it learns from new data and improves the model's accuracy. For example, the model is retrained using feedback data from Salesperson A.

[0681] Through these steps, a series of processes are implemented to provide each salesperson with the most appropriate role-play scenario, thereby maximizing the effectiveness of sales training.

[0682] Example 1

[0683] 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."

[0684] Conventional salesperson training systems have the problem of being dependent on a uniform training program, making it difficult to customize based on the characteristics and performance of each individual salesperson. This means that it is difficult to effectively improve the performance of each salesperson, and the effectiveness of training is limited. In addition, there is a lack of a mechanism for flexibly improving the training program based on feedback, which creates challenges in long-term salesperson development.

[0685] 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.

[0686] In this invention, the server includes means for collecting sales data, market data, store data, and salesperson data, means for preprocessing the collected data, means for analyzing the preprocessed data using statistical techniques and machine learning algorithms to extract features, means for generating role-play scenarios using a neural network or the like based on the extracted features, means for distributing the generated scenarios to the terminals of each salesperson, means for the salespersons to use their terminals to perform role-plays and collect feedback, and means for analyzing the collected feedback and updating the machine learning model. This makes it possible to provide a training program customized for each salesperson and enable continuous improvement based on the feedback.

[0687] "Sales Data" refers to information relating to the sales of goods and services at a store.

[0688] "Market Data" means information about external market trends and economic factors.

[0689] "Store data" refers to information regarding the number of customers visiting the store and the store's operating status.

[0690] "Salesperson Data" means information about the accomplishments and performance of individual salespersons.

[0691] "Preprocessing" refers to processing the collected data such as filling in missing values, standardizing the format, and deleting inappropriate data.

[0692] "Analysis" refers to the extraction of features and patterns from data using statistical methods and machine learning algorithms.

[0693] A "role play scenario" is a training scenario used by salespeople for practice or simulation.

[0694] A "neural network" is a type of algorithm consisting of multiple layers of artificial neurons and used for pattern recognition and machine learning.

[0695] A "terminal" is a device, such as a computer or tablet, used by a salesperson.

[0696] "Feedback" refers to opinions and results regarding the effectiveness of the scenario and areas for improvement that the salesperson enters after performing the role-play scenario.

[0697] A "machine learning model" is an algorithm that analyzes data and makes predictions and classifications based on the results.

[0698] The system of the present invention optimizes salesperson training through multiple steps: data collection, preprocessing, analysis, scenario generation, scenario distribution, role-playing, feedback collection, feedback analysis, and updating of machine learning models. This system provides role-playing scenarios customized for each salesperson, enabling effective training.

[0699] First, the server collects sales data, market data, store data, and salesperson data. Specifically, sales data is obtained from the POS system, market data from an external API, store data from customer count sensors, and salesperson data from the HR system. These data are integrated and stored in a database on the server.

[0700] The server then pre-processes the collected data, which includes data cleansing (filling in missing data, removing irrelevant data), standardizing formats (for example, standardizing date formats), and removing duplicate data, making the data suitable for analysis.

[0701] Next, the server analyzes the preprocessed data. Statistical methods and machine learning algorithms are used for the analysis. Specific tools used include Python's Pandas and Scikit-learn. At this stage, data features and patterns are extracted. For example, analyzing the sales data of Salesperson A may reveal a strong sales trend during certain time periods.

[0702] After the features are extracted, the server uses a generative AI model such as a neural network to generate role-play scenarios based on these features. The scenarios include content that addresses specific weaknesses of the salesperson. For example, a scenario is generated for Salesperson A to improve his or her closing skills with customers during weak times.

[0703] The generated scenario is sent from the server to the salesperson's terminal. After the scenario is sent, the user (salesperson) uses the terminal to perform the role-play. Examples of specific prompts for the role-play include the following:

[0704] Prompt: "How can I approach a male customer in his 30s who visits my store between 3:00 PM and 5:00 PM on Sundays?"

[0705] After the user performs a role-play based on the scenario, they input feedback into the terminal, including the effectiveness of the scenario, areas for improvement, and performance.

[0706] The collected feedback is sent to a server, which analyzes it and updates the machine learning model based on the analysis results. This update improves the accuracy of the next scenario generation, making training for salespeople even more effective.

[0707] This system provides training scenarios optimized for each salesperson, maximizing the effectiveness of the training. For example, by implementing the scenarios provided by the system, Salesperson A's success rate in closing sales improved even during times when he was previously weak at dealing with customers.

[0708] In this way, the present invention improves the accuracy and effectiveness of salesperson training through a series of processes from data collection to model updating.

[0709] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0710] Step 1: Data collection

[0711] The server collects sales data, market data, store data, and salesperson data. Data collection is performed using API requests to obtain data from the POS system, external APIs, customer count sensors, and HR systems. Each piece of data is sent to the server in JSON or CSV format. The specific operation of data collection is for the server to periodically make API calls from each data source and store the obtained data in a database.

[0712] Input: Various external and internal data obtained by API requests

[0713] Output: Sales data, market data, store data, salesperson data stored in the database

[0714] Step 2: Data Preprocessing

[0715] The server preprocesses the collected data. Preprocessing includes data cleansing, format standardization, and removal of duplicate data. Missing data is filled in with the average value from the past, and inappropriate data is removed. To standardize the format, the date format is standardized to "YYYY-MM-DD." Specifically, the server uses the Python Pandas library to manipulate data frames and perform cleansing and format standardization.

[0716] Input: Raw data stored in the database (sales data, market data, store data, salesperson data)

[0717] Output: Preprocessed data (cleaned data)

[0718] Step 3: Data analysis

[0719] The server analyzes the preprocessed data. Statistical methods and machine learning algorithms are used for this analysis. Specific methods include regression analysis and clustering using Python's Scikit-learn. The server extracts each salesperson's characteristics and trends based on the salesperson data. For example, it can discover the sales patterns of salesperson A, who is strong during specific time periods.

[0720] Input: Preprocessed data (cleaned data)

[0721] Output: Analyzed feature data (statistical results and machine learning model output)

[0722] Step 4: Scenario generation

[0723] The server generates role-play scenarios based on the analyzed feature data. Using neural networks and generative AI models, optimal scenarios are created for each salesperson. For example, a scenario is generated for Salesperson A to improve his closing skills with customers during a specific time period. The generated scenarios are converted into natural-sounding sentences using natural language generation (NLG) technology.

[0724] Input: Analyzed feature data (features for each salesperson)

[0725] Output: Generated role-play scenario (text format)

[0726] Step 5: Scenario distribution and execution

[0727] The server distributes the generated scenario to each salesperson's device. The salesperson (user) uses the device to execute the scenario. For example, Salesperson A performs a customer service simulation while reading the prompts displayed on the device. Push notifications and APIs are used to distribute the scenario.

[0728] Input: Generated role-play scenario (text format)

[0729] Output: Scenario delivered to the terminal (text displayed on the salesperson's terminal)

[0730] Step 6: Gather feedback

[0731] After the user (salesperson) performs the role-play scenario, they input their feedback into the terminal. The feedback includes the effectiveness of the scenario, areas for improvement, and performance. The terminal then sends this feedback data to the server.

[0732] Input: Feedback entered by the salesperson into the terminal (text format)

[0733] Output: Feedback data sent to the server (JSON or CSV format)

[0734] Step 7: Analyze feedback and update the model

[0735] The server analyzes the collected feedback data and updates the machine learning model. For analysis, the server uses a Python library to retrain the model parameters based on the feedback data, which makes the next scenario generation more accurate.

[0736] Input: Feedback data stored on the server (text, JSON, CSV)

[0737] Output: Updated machine learning model (improved AI model)

[0738] This series of processes systematically supports salespeople in improving their performance, and continuous learning effects can be expected.

[0739] (Application example 1)

[0740] 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."

[0741] Aiming to improve the performance of salespeople in brick-and-mortar stores, there is a need to effectively improve sales skills, especially closing skills, by providing training scenarios tailored to each individual salesperson. Traditional training methods have struggled to effectively incorporate feedback and ensure continuous improvement. Furthermore, there are limited means of evaluating training content in real time using technologies such as voice recognition. Therefore, a system that can accurately address each salesperson's weaknesses is needed.

[0742] 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.

[0743] In this invention, the server includes means for collecting sales data, market data, store data, and salesperson data, means for preprocessing the collected data, means for analyzing the preprocessed data and extracting features, means for generating role-play scenarios based on the extracted features, means for distributing the generated scenarios to terminals of each salesperson, means for the salesperson to perform role-plays based on the scenarios and collect feedback, means for updating a machine learning model based on the collected feedback, means for generating scenarios for the salesperson to improve their closing skills for a specific product, means for the terminal to automatically evaluate responses in the role-plays using a voice recognition function, and means for using the feedback data to reflect the feedback data in the generation of the next scenario. This makes it possible to generate and evaluate training scenarios tailored to individual salespersons and reflect the feedback.

[0744] "Sales data" is information relating to product sales, including sales quantity, sales amount, sales date, sales time, and the like.

[0745] "Market data" is information about the current market situation, such as competitor trends, fluctuations in product prices, and consumer purchasing intentions.

[0746] "Store data" is information about a specific store, including the number of customers, inventory status, store layout, business hours, and the like.

[0747] "Salesperson data" is information about each salesperson, including past sales performance, customer service skills, working hours, and training history.

[0748] "Preprocessing" refers to preparing the collected raw data in a format that can be analyzed, and includes steps such as data cleansing, format standardization, and missing value imputation.

[0749] "Feature extraction" refers to finding important patterns and trends through data analysis, using statistical analysis and machine learning algorithms to reveal the essential parts of the data.

[0750] A "role-play scenario" is a training scenario that allows salespeople to improve their skills by interacting with virtual customers, and includes specific customer service situations and troubleshooting methods.

[0751] "Terminal" refers to the device used by the salesperson, which may be a smartphone, tablet, or other portable computing device.

[0752] "Feedback" refers to salespeople providing their opinions on the effectiveness of the scenario and areas for improvement after the role-play, and this information is collected and used as valuable data to improve the system.

[0753] A "machine learning model" is an algorithm that generates patterns and predictions based on past data. It is used to update the model based on collected feedback and improve the quality of training scenarios.

[0754] The "voice recognition function" is a technology that allows the terminal to analyze the salesperson's voice in real time and process the spoken content as text data, and is used to automatically evaluate the responses in role-playing.

[0755] "Closing skills" are techniques for getting customers to make a purchase decision at the final stage of sales negotiations or customer service, and are important abilities that include persuasiveness and communication skills.

[0756] The system according to the present invention has the function of collecting sales data, market data, store data, and salesperson data, and then generating, distributing, and evaluating role-play scenarios based on these data. The specific processing of this system is explained below in natural language.

[0757] First, the server collects data from external and internal sources, including store sales data, foot traffic, market data, and each salesperson's past performance data, and stores the collected data in a database.

[0758] The server then preprocesses the stored data, which includes cleansing the data, standardizing the format, and removing duplicate data. For example, it imputes missing data and removes irrelevant data. At this stage, the preprocessed data is prepared for analysis.

[0759] After the preprocessing is complete, the server performs data analysis. This analysis is carried out using statistical methods and machine learning algorithms to extract data characteristics. For example, analyzing the data of Salesperson A reveals that he tends to be strong at certain times of the day and weak at other times of the day.

[0760] Next, the server generates role-play scenarios optimized for each salesperson based on the extracted features. Neural networks and other AI techniques are used to generate these scenarios. For example, a scenario is generated for Salesperson A to improve his closing skills with customers during weak times. The prompt might be something like, "Salesperson A is having trouble closing sales for a specific product. Please generate a role-play scenario to improve his skills in promoting this product."

[0761] The generated scenario is distributed by the server to the target salesperson's terminal. Once the scenario is distributed, the salesperson uses the terminal to perform role-playing. For example, Salesperson A receives a scenario to practice closing a specific product, and then interacts with a virtual customer based on that scenario.

[0762] After the role-play, the salesperson enters feedback into the terminal, including the effectiveness of the scenario, areas for improvement, and performance. The terminal may also automatically evaluate the responses during the role-play using voice recognition.

[0763] The collected feedback is sent to the server. The server analyzes the feedback data and updates the machine learning model. This update improves the accuracy of the next scenario generation. For example, by incorporating feedback from Salesperson A, a more effective next scenario is generated.

[0764] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0765] Step 1:

[0766] First, the server collects data from external and internal sources, including store sales data, customer traffic, market data, and each salesperson's past performance data. The input is data from the sales system, customer management system, and market information system, and the output is stored in a database.

[0767] Step 2:

[0768] Next, the server preprocesses the stored data. This includes data cleansing, formatting standardization, and deleting duplicate data. The input is the raw data collected in step 1, and the output is data in a clean, analyzable format. Specifically, missing values ​​are imputed, inappropriate data is deleted, and the format is standardized.

[0769] Step 3:

[0770] The server then performs data analysis based on the preprocessed data. Statistical methods and machine learning algorithms are used for the analysis. For example, based on the data of Salesperson A, strengths and weaknesses in specific time periods can be extracted. The input is the preprocessed data, and the output is an index showing the characteristics of the data.

[0771] Step 4:

[0772] Next, the server generates role-play scenarios optimized for each salesperson based on the extracted features. Neural networks and other AI techniques are used to generate the scenarios. For example, a prompt such as "Salesperson A is struggling to close a specific product. Please generate a role-play scenario to help him improve his skills in promoting this product" can be input, generating a scenario focused on specific closing skills. The output is a specific role-play scenario.

[0773] Step 5:

[0774] The generated scenario is delivered by the server to the target salesperson's terminal. The input is the generated role-play scenario, and the output is the delivered scenario. The salesperson can receive the role-play scenario on their terminal.

[0775] Step 6:

[0776] After the scenario is delivered, the salesperson uses the terminal to perform the role-play. Following the scenario on the terminal, the salesperson interacts with a virtual customer. The input is the delivered scenario, and the output is the result of the role-play.

[0777] Step 7:

[0778] After the role-play, the salesperson enters their feedback into the terminal. The feedback includes the effectiveness of the scenario, areas for improvement, and performance. The input is the results of the role-play and the salesperson's opinions, and the output is structured feedback data.

[0779] Step 8:

[0780] The collected feedback is sent to the server, which analyzes the feedback data and updates the machine learning model. The input is the feedback data, and the output is an updated machine learning model. This improves the accuracy of the next scenario generation.

[0781] Through these steps, a series of processes are carried out, providing each salesperson with a role-play scenario that is optimized for them.

[0782] 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.

[0783] This invention relates to a system that collects and analyzes sales data, market data, store data, and salesperson data, and generates role-play scenarios based on these data. Furthermore, by combining this with an emotion engine that recognizes the emotions of users (salespersons), more effective training can be achieved.

[0784] First, the server collects sales data, market data, store data, and salesperson data from external and internal sources, using APIs and database queries to retrieve various data and store it in a database.

[0785] Next, the server preprocesses the collected data. Since the collected data is often incomplete, missing data is filled in, duplicate data is removed, and the format is standardized. This prepares the data for analysis.

[0786] The server then performs data analysis based on the preprocessed data. Statistical methods and machine learning models are used to extract the characteristics of each salesperson and store. For example, patterns can be discovered, such as strengths on certain days of the week or at certain times of the day, but weaknesses on other days or at certain times of the day.

[0787] The server then generates role-play scenarios based on the extracted features. Using neural networks and other AI techniques, scenarios optimized for each salesperson and store are generated. For example, for Salesperson A, a scenario is created to strengthen closing skills to overcome weak performance.

[0788] The generated scenario is delivered from the server to the target salesperson's terminal. The scenario is sent to the terminal using notifications or file transfers. The user (salesperson) then performs a role-play based on the delivered scenario.

[0789] When the user is performing a role-play, the emotion engine installed in the device analyzes the salesperson's facial expressions and tone of voice to recognize their emotions. For example, if Salesperson A is feeling stressed, this information will be detected by the emotion engine.

[0790] The emotional information analyzed by the emotion engine is sent to the server. The server then dynamically adjusts the difficulty and content of the role-play scenario based on this information. For example, if Salesperson A feels stressed by a difficult scenario, the difficulty of the scenario can be lowered.

[0791] After the role-playing is completed, the user inputs feedback into the terminal. The feedback includes the results of the role-playing, impressions, and areas for improvement. The server collects this feedback and stores it in a database.

[0792] The collected feedback is then analyzed by the server again to update the machine learning model, which will enable it to provide more accurate and effective scenarios the next time it generates them.

[0793] In this way, by providing customized role-play scenarios that take into account the emotions of salespeople, it is expected that the effectiveness of training will be maximized and salespeople's performance will improve.

[0794] The processing flow will be explained below.

[0795] MODE FOR CARRYING OUT THE INVENTION

[0796] Step 1:

[0797] The server collects sales data, market data, store data, and salesperson data from external and internal sources. Specifically, it retrieves various data using APIs and database queries and stores it in a database. For example, store sales data and market trend data can be retrieved through an API and stored in the appropriate database tables.

[0798] Step 2:

[0799] The server preprocesses the collected data. Specific operations include filling in missing data, deleting duplicate data, and standardizing formats. For example, if there are gaps in sales data, it fills in the gaps with the average value from the past, and if the data format is inconsistent, it converts it into a unified format.

[0800] Step 3:

[0801] The server then performs data analysis using the preprocessed data. Specifically, statistical methods and machine learning models are used to extract patterns and features from the data. For example, analysis can be performed to discover a trend in which Salesperson A's performance is weak during certain time periods.

[0802] Step 4:

[0803] The server generates role-play scenarios based on the extracted features. As a specific example, neural networks and AI technology are used to create scenarios that strengthen salespeople's weaknesses. For example, a role-play scenario is generated to strengthen Salesperson A's closing skills when dealing with customers during weekdays.

[0804] Step 5:

[0805] The server generates a role-play scenario and distributes it to the salesperson's device. Specifically, it uses notification and file transfer functions to send the scenario to the salesperson's tablet or smartphone.

[0806] Step 6:

[0807] The user (salesperson) uses the terminal to perform role-playing. Specifically, the user follows a scenario displayed on the terminal to simulate customer interactions. For example, a role-playing session is conducted in which Salesperson A tries out a new closing technique.

[0808] Step 7:

[0809] During the role-play, the device's built-in emotion engine recognizes the user's emotions. For example, it uses a camera and microphone to analyze the salesperson's facial expressions and tone of voice, collecting emotional data such as stress and satisfaction.

[0810] Step 8:

[0811] The server receives the emotion data sent from the emotion engine and dynamically adjusts the difficulty and content of the role-play scenario. For example, if Salesperson A is feeling stressed, the server adjusts the difficulty of the next scenario based on this information.

[0812] Step 9:

[0813] The user enters feedback on the role-play into the terminal. Specifically, after the role-play is completed, the user enters the results, impressions, and points for improvement into the feedback form and submits it. For example, Salesperson A enters feedback such as, "The closing techniques in the scenario were effective."

[0814] Step 10:

[0815] The server analyzes the collected feedback and stores it in a database. Specifically, it receives and analyzes the feedback data to update the machine learning model. This data is reflected in the next scenario generation. For example, the feedback from Salesperson A is used to learn how to create a more precise scenario for the next time.

[0816] This process maximizes the effectiveness of training by providing each salesperson with tailored role-play scenarios.

[0817] Example 2

[0818] 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."

[0819] Conventional salesperson training systems have difficulty providing real-time feedback on salespersons' actual performance and emotions, making it impossible to provide training scenarios optimized for each individual salesperson. Furthermore, there was no system in place to properly collect and analyze training feedback and reflect it in the next training session. As a result, the effectiveness of training was limited.

[0820] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0821] In this invention, the server includes means for collecting sales data, market data, store data, and salesperson data, means for preprocessing the collected data, means for analyzing the preprocessed data and extracting features, means for generating role-play scenarios based on the extracted features using a neural network or other generative AI model, means for distributing the generated scenarios to each salesperson's terminal, means for the salesperson to perform role-plays based on the scenarios and for recognizing emotions using an emotion engine of the terminal, means for dynamically adjusting the difficulty and content of the role-play scenarios based on the emotion information, means for collecting feedback after the scenarios are performed, and means for updating the machine learning model based on the collected feedback. This makes it possible to provide effective training scenarios based on the salesperson's actual performance and emotions and to reflect the feedback in real time.

[0822] "Sales Data" refers to information related to the sale of goods or services, including sales quantity, sales amount, sales date and time, customer information, etc.

[0823] "Market data" refers to information related to market conditions, including the balance of supply and demand, price fluctuations, competitive conditions, and economic indicators.

[0824] "Store data" refers to information related to a specific store, including the store's location, size, business hours, staff information, sales performance, etc.

[0825] "Salesperson Data" means information related to individual salespersons, including their skills, experience, ratings, and performance.

[0826] "Preprocessing" refers to tasks such as data cleaning and format conversion to convert collected data into an analyzable format.

[0827] "Feature extraction" refers to the process of finding important patterns and relationships in pre-processed data.

[0828] "Generative AI models" refer to technologies that use neural networks and other artificial intelligence algorithms to generate new information and scenarios from data.

[0829] A "role-play scenario" refers to a scenario created for the purpose of training salespeople to practice sales activities in specific situations.

[0830] "Terminal" refers to the digital device used by the salesperson, including PCs, tablets, smartphones, etc.

[0831] An "emotion engine" refers to software or hardware that analyzes and recognizes emotions from a salesperson's facial expressions, tone of voice, etc.

[0832] "Feedback" refers to comments and evaluations provided by salespeople after the role-play, including information on the effectiveness of the training and areas for improvement.

[0833] A "machine learning model" is an algorithm that makes predictions or classifications based on data and is trained using training data.

[0834] The present invention relates to a system that collects and analyzes sales data, market data, store data, and salesperson data, and generates role-play scenarios based on these data. Furthermore, by combining this system with an emotion engine that recognizes the emotions of users (salespersons), it is possible to achieve more effective training. Specific embodiments are described below.

[0835] First, the server collects sales data, market data, store data, and salesperson data from external and internal sources. To do this, it uses APIs and database queries to retrieve various data and stores it in a database. Specifically, it uses RESTful APIs and SQL queries. For example, the server retrieves market data from "https: / / api.example.com / marketdata" and retrieves sales data with the SQL query "SELECT FROM sales WHERE date='2023-10-01'".

[0836] Next, the data collected by the server is preprocessed. As the data is often incomplete as it is, missing data is complemented, duplicate data is deleted, and the format is standardized. Specifically, using the Python Pandas library, duplicate data is deleted using the ".drop_duplicates()" method, and missing data is complemented using the ".fillna(method='ffill')" method.

[0837] The server performs data analysis based on the preprocessed data. It uses statistical methods and machine learning models (e.g., regression analysis, clustering) to extract the characteristics of each salesperson and store. For example, the server uses the Scikit-learn library to analyze the relationship between sales data and days of the week using the "LinearRegression()" model.

[0838] Next, the server generates role-play scenarios based on the analysis results. It uses neural networks or generative AI models (e.g., GPT-3, BERT) to generate scenarios optimized for each salesperson and store. Specifically, it uses the TensorFlow library to create a neural network model with "tf.keras.models.Sequential()" to generate a scenario specifically for Salesperson A.

[0839] The generated scenario is distributed from the server to the terminal of the target salesperson. The scenario is sent to the terminal using notification or file transfer. Specifically, the file is transferred using an HTTP request or FTP. For example, the server uses "requests.post(url, data=scenario data)" to send the scenario to Salesperson A's terminal.

[0840] The user (salesperson) performs a role-play based on the delivered scenario. During the role-play, the device's built-in emotion engine (e.g., OpenFace or IBM Watson Tone Analyzer) analyzes the salesperson's facial expressions and tone of voice to recognize their emotions. For example, the device uses the OpenFace library to analyze emotions such as "stress" or "confidence" from the user's facial expression data.

[0841] The emotional information analyzed by the emotion engine is sent to the server. The server uses this information to dynamically adjust the difficulty and content of the role-play scenario. Specifically, it uses a reinforcement learning model to change the difficulty of the scenario in real time. For example, if it receives information that Salesperson A is feeling stressed, it will set the difficulty of the scenario to a lower level.

[0842] After the role-play is completed, the user enters feedback into the device. The feedback includes the results of the role-play, their impressions, and areas for improvement. Specifically, opinions are collected using Google Forms or a feedback form developed in-house. For example, the user may enter their impression that "my closing skills have improved" into the form displayed on the device.

[0843] The server analyzes the collected feedback and updates the machine learning model. The feedback data is used to improve the model's accuracy and is reflected in the next scenario generation. Specifically, the feedback data is cleaned using the Pandas library and the model is retrained using Scikit-learn.

[0844] Prompt Sentence Examples

[0845] "You are the server of a system that collects and analyzes sales data, market data, store data, and salesperson data, and generates role-play scenarios based on this data. Specifically, please perform the following tasks:

[0846] 1. Collect data using API queries or SQL queries and store it in a database.

[0847] 2. Preprocess the data using the Pandas library.

[0848] 3. Analyze the data using a machine learning model to extract the characteristics of each salesperson.

[0849] 4. Generate role-play scenarios using TensorFlow.

[0850] 5. The scenario is delivered to the user's device via an HTTP request.

[0851] 6. The user performs a role-play based on the scenario, and the device's emotion engine recognizes emotions.

[0852] 7. Adjust the difficulty of the scenario based on emotional data.

[0853] 8. Collect user feedback and incorporate it into the next scenario generation.

[0854] For example, if Salesperson A is feeling stressed, adjust the scenario to be less difficult and use the feedback data to update the machine learning model for effective training.

[0855] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0856] Step 1: Data collection

[0857] The server obtains sales data, market data, store data, and salesperson data from external and internal sources.

[0858] Input: API query, SQL query

[0859] Output: Collected data (sales data, market data, store data, salesperson data)

[0860] The server retrieves market information using a RESTful API and pulls store information from an internal database using SQL queries. For example, it retrieves market data from "https: / / api.example.com / marketdata" and retrieves sales data with the SQL query "SELECT FROM sales WHERE date='2023-10-01'".

[0861] Step 2: Data Preprocessing

[0862] The server preprocesses the collected data.

[0863] Input: Collected data

[0864] Output: Preprocessed data

[0865] The server uses the Python Pandas library to clean the data. Specifically, it removes duplicate data using the ".drop_duplicates()" method, fills missing data using the ".fillna(method='ffill')" method, and standardizes the data format.

[0866] Step 3: Data analysis

[0867] The server performs analysis based on the preprocessed data and extracts features.

[0868] Input: Preprocessed data

[0869] Output: Extracted feature data

[0870] The server uses the Scikit-learn library to analyze the data using statistical methods and machine learning models (e.g., regression analysis, clustering). Specifically, it uses the "LinearRegression()" model to analyze the relationship between sales data and days of the week and extract features.

[0871] Step 4: Role-play scenario generation

[0872] The server generates a role-play scenario based on the extracted features.

[0873] Input: Extracted feature data

[0874] Output: Generated role-play scenario

[0875] Generate scenarios using neural networks or generative AI models (e.g., GPT-3, BERT). Specifically, use the TensorFlow library to create a neural network model with "tf.keras.models.Sequential()" to generate scenarios specifically for Salesperson A.

[0876] Step 5: Scenario Distribution

[0877] The server distributes the generated scenario to each salesperson's terminal.

[0878] Input: Generated role-play scenario

[0879] Output: Scenario sent to salesperson's terminal

[0880] The scenario is sent to the terminal using notification or file transfer. Specifically, the file is transferred using an HTTP request or FTP. For example, the server uses "requests.post(url, data=scenario data)" to send the scenario to Salesperson A's terminal.

[0881] Step 6: Role-playing and emotion recognition

[0882] The user performs role-playing based on the delivered scenario.

[0883] Input: Streamed role-play scenario

[0884] Output: Emotion data during role-playing

[0885] The device's built-in emotion engine (e.g., OpenFace or IBM Watson Tone Analyzer) analyzes the salesperson's facial expressions and tone of voice to recognize their emotions. For example, the device uses the OpenFace library to analyze emotions such as "stress" or "confidence" from the user's facial expression data.

[0886] Step 7: Scenario adjustment

[0887] The server dynamically adjusts the difficulty and content of the role-play scenario based on the emotional information obtained from the emotion engine.

[0888] Input: Emotion data

[0889] Output: A tailored role-play scenario

[0890] Use a reinforcement learning model to change the difficulty of a scenario in real time. For example, receive information that Salesperson A is feeling stressed and set the difficulty of the scenario to a lower level.

[0891] Step 8: Gather feedback

[0892] After the role-play is completed, the user inputs feedback into the terminal.

[0893] Input: User feedback

[0894] Output: Collected feedback data

[0895] The feedback includes the results of the role-play, impressions, and areas for improvement. Specifically, opinions are collected using Google Forms or an in-house developed feedback form. For example, users can enter their impressions, such as "My closing skills have improved," into the form displayed on their device.

[0896] Step 9: Feedback analysis and model updating

[0897] The server analyzes the collected feedback and updates the machine learning model.

[0898] Input: Collected feedback data

[0899] Output: An updated machine learning model

[0900] The feedback data is used to improve the accuracy of the model and reflected in the next scenario generation. Specifically, the feedback data is cleaned using the Pandas library and the model is retrained using Scikit-learn.

[0901] (Application example 2)

[0902] 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."

[0903] Conventional salesperson training systems lacked the ability to automatically generate scenarios optimized for each salesperson and dynamically adjust the difficulty and content based on their emotions. This made it difficult to provide effective training tailored to each salesperson's individual abilities and characteristics, limiting performance improvement. Furthermore, the inability to analyze salesperson emotions in real time and reflect them in the training content reduced the effectiveness of the training.

[0904] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for collecting sales data, market data, store data, and salesperson data; means for preprocessing the collected data; means for analyzing the preprocessed data and extracting features; means for generating role-play scenarios based on the extracted features; means for distributing the generated scenarios to a terminal including means for analyzing the emotions of the salesperson; means for the salesperson to perform role-plays based on the scenarios, adjusting the difficulty and content of the scenarios based on the emotion analysis results, and collecting feedback; and means for updating the machine learning model based on the collected feedback and generating the next scenario. This makes it possible to generate and adjust role-play scenarios optimized for individual salespersons, taking into account the emotions of the salespersons.

[0905] "Sales data" refers to data that indicates sales and transaction information of products related to sales activities.

[0906] "Market data" refers to data that reflects market conditions and trends, including the actions of competitors and consumer trends.

[0907] "Store data" is data related to a specific store, and includes store sales information, inventory status, customer demographics, and the like.

[0908] "Salesperson data" refers to data relating to individual salespersons, including sales performance, skill level, working hours, and the like.

[0909] "Preprocessing" refers to processes such as filling in missing values ​​in collected data, deleting duplicate data, and standardizing formats.

[0910] "Feature extraction" is the process of finding useful patterns and trends in the data being analyzed.

[0911] A "role play scenario" is a simulated sales situation scenario that a salesperson performs for learning and training purposes.

[0912] The "emotion engine" is an engine that analyzes the facial expressions and tone of voice of salespeople to recognize their emotions.

[0913] A "machine learning model" is an algorithm or system that learns from large amounts of data and performs pattern recognition and prediction.

[0914] "Feedback" refers to information such as evaluations, impressions, and areas for improvement collected from salespeople after the role-play is completed.

[0915] In this invention, a server, a terminal, and a user (salesperson) work together to realize a training system for salespersons. Specific embodiments for carrying out the invention are described below.

[0916] First, the server collects sales data, market data, store data, and salesperson data. This is done using APIs and database queries to obtain various data and store it in the database. It is also possible to obtain real-time data from external data sources.

[0917] Next, the server preprocesses the collected data, completing missing data, removing duplicate data, and standardizing the data format to prepare it for analysis. This preprocessing improves the accuracy of the analysis and enables more reliable feature extraction.

[0918] Based on the preprocessed data, the server performs data analysis and extracts characteristics of salespeople and stores. For example, it uses statistical methods and machine learning models to discover patterns such as which salespeople are effective on certain days or at certain times of the day, or which stores are effective in particular products.

[0919] After the features are extracted, the server generates role-play scenarios, leveraging neural networks and other generative AI models to generate scenarios optimized for each salesperson and store, including the skills and situations needed to improve the salesperson's performance.

[0920] The generated scenario is distributed from the server to the salesperson's device. For example, the scenario is sent to a device such as a smartphone or head-mounted display (HMD) via notification or file transfer. The salesperson then role-plays based on the distributed scenario.

[0921] At this time, the emotion engine installed in the device analyzes the salesperson's facial expressions and tone of voice to recognize their emotions. The emotion engine is responsible for analyzing in real time whether the salesperson is feeling stressed or relaxed. For example, if the server detects that the salesperson is feeling stressed, it can dynamically adjust the difficulty of the scenario based on that information.

[0922] After the role-play is completed, the user (salesperson) enters feedback into the terminal. The feedback includes the results of the role-play, impressions, and areas for improvement. This feedback is sent to the server and used to update the machine learning model. This makes it possible to provide even more accurate and effective scenarios the next time a scenario is generated.

[0923] For example, a scenario can be generated using the following prompt:

[0924] "Generate a scenario to improve closing skills for a female in her twenties who visits your store during lunchtime on a weekday. If the salesperson feels stressed, adjust the difficulty of the scenario to be less difficult."

[0925] Based on this prompt, the server can generate the optimal role-play scenario for specific conditions and skill sets.

[0926] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0927] Step 1: Data collection

[0928] The server collects sales data, market data, store data, and salesperson data using external APIs and database queries. It uses API endpoints and database connection information as input, retrieves the latest data from each data source, and stores it in the database. The output is a collection of the collected raw data.

[0929] Step 2: Data Preprocessing

[0930] The server preprocesses the collected raw data. It uses the collected raw data as input, complements missing data, removes duplicate data, and standardizes the format to make it suitable for analysis. The output is preprocessed clean data. Specifically, it finds defects in the data, complements them, and unifies inconsistent data formats.

[0931] Step 3: Data analysis

[0932] The server analyzes the preprocessed data and extracts features. It uses the preprocessed clean data as input. It uses statistical methods and machine learning models to analyze patterns of salespeople and stores. The output is feature data for each salesperson and store. Specifically, it finds patterns of salespeople who are strong on specific days of the week or at specific times of the day, or stores with high sales.

[0933] Step 4: Role-play scenario generation

[0934] The server generates role-play scenarios based on the extracted feature data. Utilizing neural networks and generative AI models, it designs scenarios optimized for each salesperson. Feature data and prompts are used as input. The output is the generated role-play scenario. Specifically, it automatically generates scenarios that strengthen the salesperson's weaknesses.

[0935] Step 5: Scenario distribution

[0936] The server distributes the generated scenario to the salesperson's terminal. It uses the generated role-play scenario as input. It sends the scenario to the terminal using notification or file transfer. The output is a notification or file that the scenario was distributed. Specifically, it sends a notification to the terminal so that the salesperson can check the received scenario.

[0937] Step 6: Sentiment Analysis

[0938] The device uses an emotion engine to analyze the salesperson's emotions during role-playing. Data on the salesperson's facial expressions and tone of voice are used as input. The emotion engine analyzes this data and recognizes emotional states such as stress or relaxation. The output is the recognized emotion data. Specifically, the device uses a camera and microphone to capture facial expressions and voice data in real time, which the emotion engine analyzes.

[0939] Step 7: Dynamically adjust the scenario

[0940] The server adjusts the difficulty and content of the scenario based on the emotional data sent from the terminal. Emotional data is used as input. If the salesperson is feeling stressed, the server adjusts the difficulty of the scenario, for example, by lowering it. The output is the adjusted scenario. Specifically, the server uses a difficulty adjustment algorithm to dynamically change the content of the scenario.

[0941] Step 8: Gather feedback

[0942] The terminal collects feedback from the salesperson after the role-play ends. The input is the feedback data entered by the user. The output is the collected feedback data. In concrete terms, the salesperson fills out a feedback form on the terminal, which is then sent to the server.

[0943] Step 9: Update the machine learning model

[0944] The server updates the machine learning model based on the collected feedback data. It uses the feedback data and past data as input. The updated model can provide more accurate scenarios the next time it generates a scenario. The output is an updated machine learning model. Specifically, it adds the new feedback data to the learning data and retrains the model.

[0945] 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.

[0946] 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.

[0947] 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.

[0948] [Fourth embodiment]

[0949] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[0950] 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.

[0951] 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).

[0952] 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.

[0953] 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.

[0954] 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).

[0955] 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.

[0956] 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.

[0957] 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.

[0958] 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.

[0959] 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.

[0960] 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.

[0961] 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."

[0962] The system according to the present invention has the function of collecting sales data, market data, store data, and salesperson data, and then generating, distributing, and evaluating role-play scenarios based on this data. The specific processing of this program and a natural language explanation are given below.

[0963] First, the server collects data from external and internal sources, including store sales data, foot traffic, market data, and each salesperson's past performance data, and stores the collected data in a database.

[0964] The server then preprocesses the stored data, which includes cleansing the data, standardizing the format, and removing duplicate data. For example, it imputes missing data and removes irrelevant data. At this stage, the preprocessed data is prepared for analysis.

[0965] After the preprocessing is complete, the server performs data analysis. This analysis is carried out using statistical methods and machine learning algorithms to extract data characteristics. For example, analyzing the data of Salesperson A reveals that he tends to be strong at certain times of the day and weak at other times of the day.

[0966] Next, the server generates role-play scenarios optimized for each salesperson based on the extracted features. This scenario generation uses neural networks and other AI techniques. For example, a scenario is generated for Salesperson A to improve his or her closing skills with customers during weak times.

[0967] The generated scenario is distributed by the server to the terminal of the target salesperson. Once the scenario is distributed, the user (salesperson) uses the terminal to perform role-playing. After the role-playing is completed, the user enters feedback into the terminal. The feedback includes the effectiveness of the scenario, areas for improvement, and performance.

[0968] The collected feedback is sent to the server. The server analyzes the feedback data and updates the machine learning model. This update improves the accuracy of the next scenario generation. For example, by incorporating feedback from Salesperson A, a more effective next scenario is generated.

[0969] In this way, the system provides each salesperson with a customized role-play scenario, maximizing the effectiveness of the training. This system is expected to improve salesperson performance.

[0970] The processing flow will be explained below.

[0971] Specific processing flow of the program

[0972] Step 1:

[0973] The server collects data from external and internal sources, specifically sales data, market data, store data, and salesperson data, using APIs and database queries. For example, it retrieves sales data by store, market trend data, and salesperson historical performance data.

[0974] Step 2:

[0975] The server preprocesses the collected data. Specifically, it complements missing data, removes duplicate data, and standardizes the format. For example, it complements missing data with average values ​​and removes inappropriate data.

[0976] Step 3:

[0977] The server analyzes the preprocessed data. Specifically, it uses statistical techniques and machine learning models to extract patterns and features from the data. For example, it can discover whether there are strong or weak trends on certain days of the week or during certain times of the day from the data of Salesperson A.

[0978] Step 4:

[0979] The server generates role-play scenarios based on the extracted characteristics. Specifically, it uses neural networks and AI technology to generate scenarios tailored to the characteristics of each salesperson and store. For example, for Salesperson A, it creates a scenario that includes closing techniques that should be strengthened during times when Salesperson A is weak.

[0980] Step 5:

[0981] The server distributes the generated role-play scenario to the device of the salesperson. Specifically, the scenario is sent to the device using notifications or file transfers. For example, the scenario is pushed to the tablet of Salesperson A.

[0982] Step 6:

[0983] The user (salesperson) uses the terminal to carry out a role-play scenario. Specifically, the user reads the displayed scenario and acts based on it. For example, Salesperson A performs a role-play according to the scenario.

[0984] Step 7:

[0985] The user enters feedback on the role-play into the terminal. Specifically, the results and impressions are entered into a form on the terminal and sent. For example, Salesperson A enters feedback such as "The closing technique in the scenario was effective."

[0986] Step 8:

[0987] The server collects the feedback and stores it in a database. Specifically, it receives and stores the feedback data sent from the terminal. For example, the feedback of salesperson A is stored in the database.

[0988] Step 9:

[0989] The server analyzes the collected feedback and updates the machine learning model. Specifically, it learns from new data and improves the model's accuracy. For example, the model is retrained using feedback data from Salesperson A.

[0990] Through these steps, a series of processes are implemented to provide each salesperson with the most appropriate role-play scenario, thereby maximizing the effectiveness of sales training.

[0991] Example 1

[0992] 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."

[0993] Conventional salesperson training systems have the problem of being dependent on a uniform training program, making it difficult to customize based on the characteristics and performance of each individual salesperson. This means that it is difficult to effectively improve the performance of each salesperson, and the effectiveness of training is limited. In addition, there is a lack of a mechanism for flexibly improving the training program based on feedback, which creates challenges in long-term salesperson development.

[0994] 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.

[0995] In this invention, the server includes means for collecting sales data, market data, store data, and salesperson data, means for preprocessing the collected data, means for analyzing the preprocessed data using statistical techniques and machine learning algorithms to extract features, means for generating role-play scenarios using a neural network or the like based on the extracted features, means for distributing the generated scenarios to the terminals of each salesperson, means for the salespersons to use their terminals to perform role-plays and collect feedback, and means for analyzing the collected feedback and updating the machine learning model. This makes it possible to provide a training program customized for each salesperson and enable continuous improvement based on the feedback.

[0996] "Sales Data" refers to information relating to the sales of goods and services at a store.

[0997] "Market Data" means information about external market trends and economic factors.

[0998] "Store data" refers to information regarding the number of customers visiting the store and the store's operating status.

[0999] "Salesperson Data" means information about the accomplishments and performance of individual salespersons.

[1000] "Preprocessing" refers to processing the collected data such as filling in missing values, standardizing the format, and deleting inappropriate data.

[1001] "Analysis" refers to the extraction of features and patterns from data using statistical methods and machine learning algorithms.

[1002] A "role play scenario" is a training scenario used by salespeople for practice or simulation.

[1003] A "neural network" is a type of algorithm consisting of multiple layers of artificial neurons and used for pattern recognition and machine learning.

[1004] A "terminal" is a device, such as a computer or tablet, used by a salesperson.

[1005] "Feedback" refers to opinions and results regarding the effectiveness of the scenario and areas for improvement that the salesperson enters after performing the role-play scenario.

[1006] A "machine learning model" is an algorithm that analyzes data and makes predictions and classifications based on the results.

[1007] The system of the present invention optimizes salesperson training through multiple steps: data collection, preprocessing, analysis, scenario generation, scenario distribution, role-playing, feedback collection, feedback analysis, and updating of machine learning models. This system provides role-playing scenarios customized for each salesperson, enabling effective training.

[1008] First, the server collects sales data, market data, store data, and salesperson data. Specifically, sales data is obtained from the POS system, market data from an external API, store data from customer count sensors, and salesperson data from the HR system. These data are integrated and stored in a database on the server.

[1009] The server then pre-processes the collected data, which includes data cleansing (filling in missing data, removing irrelevant data), standardizing formats (for example, standardizing date formats), and removing duplicate data, making the data suitable for analysis.

[1010] Next, the server analyzes the preprocessed data. Statistical methods and machine learning algorithms are used for the analysis. Specific tools used include Python's Pandas and Scikit-learn. At this stage, data features and patterns are extracted. For example, analyzing the sales data of Salesperson A may reveal a strong sales trend during certain time periods.

[1011] After the features are extracted, the server uses a generative AI model such as a neural network to generate role-play scenarios based on these features. The scenarios include content that addresses specific weaknesses of the salesperson. For example, a scenario is generated for Salesperson A to improve his or her closing skills with customers during weak times.

[1012] The generated scenario is sent from the server to the salesperson's terminal. After the scenario is sent, the user (salesperson) uses the terminal to perform the role-play. Examples of specific prompts for the role-play include the following:

[1013] Prompt: "How can I approach a male customer in his 30s who visits my store between 3:00 PM and 5:00 PM on Sundays?"

[1014] After the user performs a role-play based on the scenario, they input feedback into the terminal, including the effectiveness of the scenario, areas for improvement, and performance.

[1015] The collected feedback is sent to a server, which analyzes it and updates the machine learning model based on the analysis results. This update improves the accuracy of the next scenario generation, making training for salespeople even more effective.

[1016] This system provides training scenarios optimized for each salesperson, maximizing the effectiveness of the training. For example, by implementing the scenarios provided by the system, Salesperson A's success rate in closing sales improved even during times when he was previously weak at dealing with customers.

[1017] In this way, the present invention improves the accuracy and effectiveness of salesperson training through a series of processes from data collection to model updating.

[1018] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1019] Step 1: Data collection

[1020] The server collects sales data, market data, store data, and salesperson data. Data collection is performed using API requests to obtain data from the POS system, external APIs, customer count sensors, and HR systems. Each piece of data is sent to the server in JSON or CSV format. The specific operation of data collection is for the server to periodically make API calls from each data source and store the obtained data in a database.

[1021] Input: Various external and internal data obtained by API requests

[1022] Output: Sales data, market data, store data, salesperson data stored in the database

[1023] Step 2: Data Preprocessing

[1024] The server preprocesses the collected data. Preprocessing includes data cleansing, format standardization, and removal of duplicate data. Missing data is filled in with the average value from the past, and inappropriate data is removed. To standardize the format, the date format is standardized to "YYYY-MM-DD." Specifically, the server uses the Python Pandas library to manipulate data frames and perform cleansing and format standardization.

[1025] Input: Raw data stored in the database (sales data, market data, store data, salesperson data)

[1026] Output: Preprocessed data (cleaned data)

[1027] Step 3: Data analysis

[1028] The server analyzes the preprocessed data. Statistical methods and machine learning algorithms are used for this analysis. Specific methods include regression analysis and clustering using Python's Scikit-learn. The server extracts each salesperson's characteristics and trends based on the salesperson data. For example, it can discover the sales patterns of salesperson A, who is strong during specific time periods.

[1029] Input: Preprocessed data (cleaned data)

[1030] Output: Analyzed feature data (statistical results and machine learning model output)

[1031] Step 4: Scenario generation

[1032] The server generates role-play scenarios based on the analyzed feature data. Using neural networks and generative AI models, optimal scenarios are created for each salesperson. For example, a scenario is generated for Salesperson A to improve his closing skills with customers during a specific time period. The generated scenarios are converted into natural-sounding sentences using natural language generation (NLG) technology.

[1033] Input: Analyzed feature data (features for each salesperson)

[1034] Output: Generated role-play scenario (text format)

[1035] Step 5: Scenario distribution and execution

[1036] The server distributes the generated scenario to each salesperson's device. The salesperson (user) uses the device to execute the scenario. For example, Salesperson A performs a customer service simulation while reading the prompts displayed on the device. Push notifications and APIs are used to distribute the scenario.

[1037] Input: Generated role-play scenario (text format)

[1038] Output: Scenario delivered to the terminal (text displayed on the salesperson's terminal)

[1039] Step 6: Gather feedback

[1040] After the user (salesperson) performs the role-play scenario, they input their feedback into the terminal. The feedback includes the effectiveness of the scenario, areas for improvement, and performance. The terminal then sends this feedback data to the server.

[1041] Input: Feedback entered by the salesperson into the terminal (text format)

[1042] Output: Feedback data sent to the server (JSON or CSV format)

[1043] Step 7: Analyze feedback and update the model

[1044] The server analyzes the collected feedback data and updates the machine learning model. For analysis, the server uses a Python library to retrain the model parameters based on the feedback data, which makes the next scenario generation more accurate.

[1045] Input: Feedback data stored on the server (text, JSON, CSV)

[1046] Output: Updated machine learning model (improved AI model)

[1047] This series of processes systematically supports salespeople in improving their performance, and continuous learning effects can be expected.

[1048] (Application example 1)

[1049] 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."

[1050] Aiming to improve the performance of salespeople in brick-and-mortar stores, there is a need to effectively improve sales skills, especially closing skills, by providing training scenarios tailored to each individual salesperson. Traditional training methods have struggled to effectively incorporate feedback and ensure continuous improvement. Furthermore, there are limited means of evaluating training content in real time using technologies such as voice recognition. Therefore, a system that can accurately address each salesperson's weaknesses is needed.

[1051] 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.

[1052] In this invention, the server includes means for collecting sales data, market data, store data, and salesperson data, means for preprocessing the collected data, means for analyzing the preprocessed data and extracting features, means for generating role-play scenarios based on the extracted features, means for distributing the generated scenarios to terminals of each salesperson, means for the salesperson to perform role-plays based on the scenarios and collect feedback, means for updating a machine learning model based on the collected feedback, means for generating scenarios for the salesperson to improve their closing skills for a specific product, means for the terminal to automatically evaluate responses in the role-plays using a voice recognition function, and means for using the feedback data to reflect the feedback data in the generation of the next scenario. This makes it possible to generate and evaluate training scenarios tailored to individual salespersons and reflect the feedback.

[1053] "Sales data" is information relating to product sales, including sales quantity, sales amount, sales date, sales time, and the like.

[1054] "Market data" is information about the current market situation, such as competitor trends, fluctuations in product prices, and consumer purchasing intentions.

[1055] "Store data" is information about a specific store, including the number of customers, inventory status, store layout, business hours, and the like.

[1056] "Salesperson data" is information about each salesperson, including past sales performance, customer service skills, working hours, and training history.

[1057] "Preprocessing" refers to preparing the collected raw data in a format that can be analyzed, and includes steps such as data cleansing, format standardization, and missing value imputation.

[1058] "Feature extraction" refers to finding important patterns and trends through data analysis, using statistical analysis and machine learning algorithms to reveal the essential parts of the data.

[1059] A "role-play scenario" is a training scenario that allows salespeople to improve their skills by interacting with virtual customers, and includes specific customer service situations and troubleshooting methods.

[1060] "Terminal" refers to the device used by the salesperson, which may be a smartphone, tablet, or other portable computing device.

[1061] "Feedback" refers to salespeople providing their opinions on the effectiveness of the scenario and areas for improvement after the role-play, and this information is collected and used as valuable data to improve the system.

[1062] A "machine learning model" is an algorithm that generates patterns and predictions based on past data. It is used to update the model based on collected feedback and improve the quality of training scenarios.

[1063] The "voice recognition function" is a technology that allows the terminal to analyze the salesperson's voice in real time and process the spoken content as text data, and is used to automatically evaluate the responses in role-playing.

[1064] "Closing skills" are techniques for getting customers to make a purchase decision at the final stage of sales negotiations or customer service, and are important abilities that include persuasiveness and communication skills.

[1065] The system according to the present invention has the function of collecting sales data, market data, store data, and salesperson data, and then generating, distributing, and evaluating role-play scenarios based on these data. The specific processing of this system is explained below in natural language.

[1066] First, the server collects data from external and internal sources, including store sales data, foot traffic, market data, and each salesperson's past performance data, and stores the collected data in a database.

[1067] The server then preprocesses the stored data, which includes cleansing the data, standardizing the format, and removing duplicate data. For example, it imputes missing data and removes irrelevant data. At this stage, the preprocessed data is prepared for analysis.

[1068] After the preprocessing is complete, the server performs data analysis. This analysis is carried out using statistical methods and machine learning algorithms to extract data characteristics. For example, analyzing the data of Salesperson A reveals that he tends to be strong at certain times of the day and weak at other times of the day.

[1069] Next, the server generates role-play scenarios optimized for each salesperson based on the extracted features. Neural networks and other AI techniques are used to generate these scenarios. For example, a scenario is generated for Salesperson A to improve his closing skills with customers during weak times. The prompt might be something like, "Salesperson A is having trouble closing sales for a specific product. Please generate a role-play scenario to improve his skills in promoting this product."

[1070] The generated scenario is distributed by the server to the target salesperson's terminal. Once the scenario is distributed, the salesperson uses the terminal to perform role-playing. For example, Salesperson A receives a scenario to practice closing a specific product, and then interacts with a virtual customer based on that scenario.

[1071] After the role-play, the salesperson enters feedback into the terminal, including the effectiveness of the scenario, areas for improvement, and performance. The terminal may also automatically evaluate the responses during the role-play using voice recognition.

[1072] The collected feedback is sent to the server. The server analyzes the feedback data and updates the machine learning model. This update improves the accuracy of the next scenario generation. For example, by incorporating feedback from Salesperson A, a more effective next scenario is generated.

[1073] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1074] Step 1:

[1075] First, the server collects data from external and internal sources, including store sales data, customer traffic, market data, and each salesperson's past performance data. The input is data from the sales system, customer management system, and market information system, and the output is stored in a database.

[1076] Step 2:

[1077] Next, the server preprocesses the stored data. This includes data cleansing, formatting standardization, and deleting duplicate data. The input is the raw data collected in step 1, and the output is data in a clean, analyzable format. Specifically, missing values ​​are imputed, inappropriate data is deleted, and the format is standardized.

[1078] Step 3:

[1079] The server then performs data analysis based on the preprocessed data. Statistical methods and machine learning algorithms are used for the analysis. For example, based on the data of Salesperson A, strengths and weaknesses in specific time periods can be extracted. The input is the preprocessed data, and the output is an index showing the characteristics of the data.

[1080] Step 4:

[1081] Next, the server generates role-play scenarios optimized for each salesperson based on the extracted features. Neural networks and other AI techniques are used to generate the scenarios. For example, a prompt such as "Salesperson A is struggling to close a specific product. Please generate a role-play scenario to help him improve his skills in promoting this product" can be input, generating a scenario focused on specific closing skills. The output is a specific role-play scenario.

[1082] Step 5:

[1083] The generated scenario is delivered by the server to the target salesperson's terminal. The input is the generated role-play scenario, and the output is the delivered scenario. The salesperson can receive the role-play scenario on their terminal.

[1084] Step 6:

[1085] After the scenario is delivered, the salesperson uses the terminal to perform the role-play. Following the scenario on the terminal, the salesperson interacts with a virtual customer. The input is the delivered scenario, and the output is the result of the role-play.

[1086] Step 7:

[1087] After the role-play, the salesperson enters their feedback into the terminal. The feedback includes the effectiveness of the scenario, areas for improvement, and performance. The input is the results of the role-play and the salesperson's opinions, and the output is structured feedback data.

[1088] Step 8:

[1089] The collected feedback is sent to the server, which analyzes the feedback data and updates the machine learning model. The input is the feedback data, and the output is an updated machine learning model. This improves the accuracy of the next scenario generation.

[1090] Through these steps, a series of processes are carried out, providing each salesperson with a role-play scenario that is optimized for them.

[1091] 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.

[1092] This invention relates to a system that collects and analyzes sales data, market data, store data, and salesperson data, and generates role-play scenarios based on these data. Furthermore, by combining this with an emotion engine that recognizes the emotions of users (salespersons), more effective training can be achieved.

[1093] First, the server collects sales data, market data, store data, and salesperson data from external and internal sources, using APIs and database queries to retrieve various data and store it in a database.

[1094] Next, the server preprocesses the collected data. Since the collected data is often incomplete, missing data is filled in, duplicate data is removed, and the format is standardized. This prepares the data for analysis.

[1095] The server then performs data analysis based on the preprocessed data. Statistical methods and machine learning models are used to extract the characteristics of each salesperson and store. For example, patterns can be discovered, such as strengths on certain days of the week or at certain times of the day, but weaknesses on other days or at certain times of the day.

[1096] The server then generates role-play scenarios based on the extracted features. Using neural networks and other AI techniques, scenarios optimized for each salesperson and store are generated. For example, for Salesperson A, a scenario is created to strengthen closing skills to overcome weak performance.

[1097] The generated scenario is delivered from the server to the target salesperson's terminal. The scenario is sent to the terminal using notifications or file transfers. The user (salesperson) then performs a role-play based on the delivered scenario.

[1098] When the user is performing a role-play, the emotion engine installed in the device analyzes the salesperson's facial expressions and tone of voice to recognize their emotions. For example, if Salesperson A is feeling stressed, this information will be detected by the emotion engine.

[1099] The emotional information analyzed by the emotion engine is sent to the server. The server then dynamically adjusts the difficulty and content of the role-play scenario based on this information. For example, if Salesperson A feels stressed by a difficult scenario, the difficulty of the scenario can be lowered.

[1100] After the role-playing is completed, the user inputs feedback into the terminal. The feedback includes the results of the role-playing, impressions, and areas for improvement. The server collects this feedback and stores it in a database.

[1101] The collected feedback is then analyzed by the server again to update the machine learning model, which will enable it to provide more accurate and effective scenarios the next time it generates them.

[1102] In this way, by providing customized role-play scenarios that take into account the emotions of salespeople, it is expected that the effectiveness of training will be maximized and salespeople's performance will improve.

[1103] The processing flow will be explained below.

[1104] MODE FOR CARRYING OUT THE INVENTION

[1105] Step 1:

[1106] The server collects sales data, market data, store data, and salesperson data from external and internal sources. Specifically, it retrieves various data using APIs and database queries and stores it in a database. For example, store sales data and market trend data can be retrieved through an API and stored in the appropriate database tables.

[1107] Step 2:

[1108] The server preprocesses the collected data. Specific operations include filling in missing data, deleting duplicate data, and standardizing formats. For example, if there are gaps in sales data, it fills in the gaps with the average value from the past, and if the data format is inconsistent, it converts it into a unified format.

[1109] Step 3:

[1110] The server then performs data analysis using the preprocessed data. Specifically, statistical methods and machine learning models are used to extract patterns and features from the data. For example, analysis can be performed to discover a trend in which Salesperson A's performance is weak during certain time periods.

[1111] Step 4:

[1112] The server generates role-play scenarios based on the extracted features. As a specific example, neural networks and AI technology are used to create scenarios that strengthen salespeople's weaknesses. For example, a role-play scenario is generated to strengthen Salesperson A's closing skills when dealing with customers during weekdays.

[1113] Step 5:

[1114] The server generates a role-play scenario and distributes it to the salesperson's device. Specifically, it uses notification and file transfer functions to send the scenario to the salesperson's tablet or smartphone.

[1115] Step 6:

[1116] The user (salesperson) uses the terminal to perform role-playing. Specifically, the user follows a scenario displayed on the terminal to simulate customer interactions. For example, a role-playing session is conducted in which Salesperson A tries out a new closing technique.

[1117] Step 7:

[1118] During the role-play, the device's built-in emotion engine recognizes the user's emotions. For example, it uses a camera and microphone to analyze the salesperson's facial expressions and tone of voice, collecting emotional data such as stress and satisfaction.

[1119] Step 8:

[1120] The server receives the emotion data sent from the emotion engine and dynamically adjusts the difficulty and content of the role-play scenario. For example, if Salesperson A is feeling stressed, the server adjusts the difficulty of the next scenario based on this information.

[1121] Step 9:

[1122] The user enters feedback on the role-play into the terminal. Specifically, after the role-play is completed, the user enters the results, impressions, and points for improvement into the feedback form and submits it. For example, Salesperson A enters feedback such as, "The closing techniques in the scenario were effective."

[1123] Step 10:

[1124] The server analyzes the collected feedback and stores it in a database. Specifically, it receives and analyzes the feedback data to update the machine learning model. This data is reflected in the next scenario generation. For example, the feedback from Salesperson A is used to learn how to create a more precise scenario for the next time.

[1125] This process maximizes the effectiveness of training by providing each salesperson with tailored role-play scenarios.

[1126] Example 2

[1127] 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."

[1128] Conventional salesperson training systems have difficulty providing real-time feedback on salespersons' actual performance and emotions, making it impossible to provide training scenarios optimized for each individual salesperson. Furthermore, there was no system in place to properly collect and analyze training feedback and reflect it in the next training session. As a result, the effectiveness of training was limited.

[1129] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1130] In this invention, the server includes means for collecting sales data, market data, store data, and salesperson data, means for preprocessing the collected data, means for analyzing the preprocessed data and extracting features, means for generating role-play scenarios based on the extracted features using a neural network or other generative AI model, means for distributing the generated scenarios to each salesperson's terminal, means for the salesperson to perform role-plays based on the scenarios and for recognizing emotions using an emotion engine of the terminal, means for dynamically adjusting the difficulty and content of the role-play scenarios based on the emotion information, means for collecting feedback after the scenarios are performed, and means for updating the machine learning model based on the collected feedback. This makes it possible to provide effective training scenarios based on the salesperson's actual performance and emotions and to reflect the feedback in real time.

[1131] "Sales Data" refers to information related to the sale of goods or services, including sales quantity, sales amount, sales date and time, customer information, etc.

[1132] "Market data" refers to information related to market conditions, including the balance of supply and demand, price fluctuations, competitive conditions, and economic indicators.

[1133] "Store data" refers to information related to a specific store, including the store's location, size, business hours, staff information, sales performance, etc.

[1134] "Salesperson Data" means information related to individual salespersons, including their skills, experience, ratings, and performance.

[1135] "Preprocessing" refers to tasks such as data cleaning and format conversion to convert collected data into an analyzable format.

[1136] "Feature extraction" refers to the process of finding important patterns and relationships in pre-processed data.

[1137] "Generative AI models" refer to technologies that use neural networks and other artificial intelligence algorithms to generate new information and scenarios from data.

[1138] A "role-play scenario" refers to a scenario created for the purpose of training salespeople to practice sales activities in specific situations.

[1139] "Terminal" refers to the digital device used by the salesperson, including PCs, tablets, smartphones, etc.

[1140] An "emotion engine" refers to software or hardware that analyzes and recognizes emotions from a salesperson's facial expressions, tone of voice, etc.

[1141] "Feedback" refers to comments and evaluations provided by salespeople after the role-play, including information on the effectiveness of the training and areas for improvement.

[1142] A "machine learning model" is an algorithm that makes predictions or classifications based on data and is trained using training data.

[1143] The present invention relates to a system that collects and analyzes sales data, market data, store data, and salesperson data, and generates role-play scenarios based on these data. Furthermore, by combining this system with an emotion engine that recognizes the emotions of users (salespersons), it is possible to achieve more effective training. Specific embodiments are described below.

[1144] First, the server collects sales data, market data, store data, and salesperson data from external and internal sources. To do this, it uses APIs and database queries to retrieve various data and stores it in a database. Specifically, it uses RESTful APIs and SQL queries. For example, the server retrieves market data from "https: / / api.example.com / marketdata" and retrieves sales data with the SQL query "SELECT FROM sales WHERE date='2023-10-01'".

[1145] Next, the data collected by the server is preprocessed. As the data is often incomplete as it is, missing data is complemented, duplicate data is deleted, and the format is standardized. Specifically, using the Python Pandas library, duplicate data is deleted using the ".drop_duplicates()" method, and missing data is complemented using the ".fillna(method='ffill')" method.

[1146] The server performs data analysis based on the preprocessed data. It uses statistical methods and machine learning models (e.g., regression analysis, clustering) to extract the characteristics of each salesperson and store. For example, the server uses the Scikit-learn library to analyze the relationship between sales data and days of the week using the "LinearRegression()" model.

[1147] Next, the server generates role-play scenarios based on the analysis results. It uses neural networks or generative AI models (e.g., GPT-3, BERT) to generate scenarios optimized for each salesperson and store. Specifically, it uses the TensorFlow library to create a neural network model with "tf.keras.models.Sequential()" to generate a scenario specifically for Salesperson A.

[1148] The generated scenario is distributed from the server to the terminal of the target salesperson. The scenario is sent to the terminal using notification or file transfer. Specifically, the file is transferred using an HTTP request or FTP. For example, the server uses "requests.post(url, data=scenario data)" to send the scenario to Salesperson A's terminal.

[1149] The user (salesperson) performs a role-play based on the delivered scenario. During the role-play, the device's built-in emotion engine (e.g., OpenFace or IBM Watson Tone Analyzer) analyzes the salesperson's facial expressions and tone of voice to recognize their emotions. For example, the device uses the OpenFace library to analyze emotions such as "stress" or "confidence" from the user's facial expression data.

[1150] The emotional information analyzed by the emotion engine is sent to the server. The server uses this information to dynamically adjust the difficulty and content of the role-play scenario. Specifically, it uses a reinforcement learning model to change the difficulty of the scenario in real time. For example, if it receives information that Salesperson A is feeling stressed, it will set the difficulty of the scenario to a lower level.

[1151] After the role-play is completed, the user enters feedback into the device. The feedback includes the results of the role-play, their impressions, and areas for improvement. Specifically, opinions are collected using Google Forms or a feedback form developed in-house. For example, the user may enter their impression that "my closing skills have improved" into the form displayed on the device.

[1152] The server analyzes the collected feedback and updates the machine learning model. The feedback data is used to improve the model's accuracy and is reflected in the next scenario generation. Specifically, the feedback data is cleaned using the Pandas library and the model is retrained using Scikit-learn.

[1153] Prompt Sentence Examples

[1154] "You are the server of a system that collects and analyzes sales data, market data, store data, and salesperson data, and generates role-play scenarios based on this data. Specifically, please perform the following tasks:

[1155] 1. Collect data using API queries or SQL queries and store it in a database.

[1156] 2. Preprocess the data using the Pandas library.

[1157] 3. Analyze the data using a machine learning model to extract the characteristics of each salesperson.

[1158] 4. Generate role-play scenarios using TensorFlow.

[1159] 5. The scenario is delivered to the user's device via an HTTP request.

[1160] 6. The user performs a role-play based on the scenario, and the device's emotion engine recognizes emotions.

[1161] 7. Adjust the difficulty of the scenario based on emotional data.

[1162] 8. Collect user feedback and incorporate it into the next scenario generation.

[1163] For example, if Salesperson A is feeling stressed, adjust the scenario to be less difficult and use the feedback data to update the machine learning model for effective training.

[1164] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1165] Step 1: Data collection

[1166] The server obtains sales data, market data, store data, and salesperson data from external and internal sources.

[1167] Input: API query, SQL query

[1168] Output: Collected data (sales data, market data, store data, salesperson data)

[1169] The server retrieves market information using a RESTful API and pulls store information from an internal database using SQL queries. For example, it retrieves market data from "https: / / api.example.com / marketdata" and retrieves sales data with the SQL query "SELECT FROM sales WHERE date='2023-10-01'".

[1170] Step 2: Data Preprocessing

[1171] The server preprocesses the collected data.

[1172] Input: Collected data

[1173] Output: Preprocessed data

[1174] The server uses the Python Pandas library to clean the data. Specifically, it removes duplicate data using the ".drop_duplicates()" method, fills missing data using the ".fillna(method='ffill')" method, and standardizes the data format.

[1175] Step 3: Data analysis

[1176] The server performs analysis based on the preprocessed data and extracts features.

[1177] Input: Preprocessed data

[1178] Output: Extracted feature data

[1179] The server uses the Scikit-learn library to analyze the data using statistical methods and machine learning models (e.g., regression analysis, clustering). Specifically, it uses the "LinearRegression()" model to analyze the relationship between sales data and days of the week and extract features.

[1180] Step 4: Role-play scenario generation

[1181] The server generates a role-play scenario based on the extracted features.

[1182] Input: Extracted feature data

[1183] Output: Generated role-play scenario

[1184] Generate scenarios using neural networks or generative AI models (e.g., GPT-3, BERT). Specifically, use the TensorFlow library to create a neural network model with "tf.keras.models.Sequential()" to generate scenarios specifically for Salesperson A.

[1185] Step 5: Scenario Distribution

[1186] The server distributes the generated scenario to each salesperson's terminal.

[1187] Input: Generated role-play scenario

[1188] Output: Scenario sent to salesperson's terminal

[1189] The scenario is sent to the terminal using notification or file transfer. Specifically, the file is transferred using an HTTP request or FTP. For example, the server uses "requests.post(url, data=scenario data)" to send the scenario to Salesperson A's terminal.

[1190] Step 6: Role-playing and emotion recognition

[1191] The user performs role-playing based on the delivered scenario.

[1192] Input: Streamed role-play scenario

[1193] Output: Emotion data during role-playing

[1194] The device's built-in emotion engine (e.g., OpenFace or IBM Watson Tone Analyzer) analyzes the salesperson's facial expressions and tone of voice to recognize their emotions. For example, the device uses the OpenFace library to analyze emotions such as "stress" or "confidence" from the user's facial expression data.

[1195] Step 7: Scenario adjustment

[1196] The server dynamically adjusts the difficulty and content of the role-play scenario based on the emotional information obtained from the emotion engine.

[1197] Input: Emotion data

[1198] Output: A tailored role-play scenario

[1199] Use a reinforcement learning model to change the difficulty of a scenario in real time. For example, receive information that Salesperson A is feeling stressed and set the difficulty of the scenario to a lower level.

[1200] Step 8: Gather feedback

[1201] After the role-play is completed, the user inputs feedback into the terminal.

[1202] Input: User feedback

[1203] Output: Collected feedback data

[1204] The feedback includes the results of the role-play, impressions, and areas for improvement. Specifically, opinions are collected using Google Forms or an in-house developed feedback form. For example, users can enter their impressions, such as "My closing skills have improved," into the form displayed on their device.

[1205] Step 9: Feedback analysis and model updating

[1206] The server analyzes the collected feedback and updates the machine learning model.

[1207] Input: Collected feedback data

[1208] Output: An updated machine learning model

[1209] The feedback data is used to improve the accuracy of the model and reflected in the next scenario generation. Specifically, the feedback data is cleaned using the Pandas library and the model is retrained using Scikit-learn.

[1210] (Application example 2)

[1211] 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."

[1212] Conventional salesperson training systems lacked the ability to automatically generate scenarios optimized for each salesperson and dynamically adjust the difficulty and content based on their emotions. This made it difficult to provide effective training tailored to each salesperson's individual abilities and characteristics, limiting performance improvement. Furthermore, the inability to analyze salesperson emotions in real time and reflect them in the training content reduced the effectiveness of the training.

[1213] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for collecting sales data, market data, store data, and salesperson data; means for preprocessing the collected data; means for analyzing the preprocessed data and extracting features; means for generating role-play scenarios based on the extracted features; means for distributing the generated scenarios to a terminal including means for analyzing the emotions of the salesperson; means for the salesperson to perform role-plays based on the scenarios, adjusting the difficulty and content of the scenarios based on the emotion analysis results, and collecting feedback; and means for updating the machine learning model based on the collected feedback and generating the next scenario. This makes it possible to generate and adjust role-play scenarios optimized for individual salespersons, taking into account the emotions of the salespersons.

[1214] "Sales data" refers to data that indicates sales and transaction information of products related to sales activities.

[1215] "Market data" refers to data that reflects market conditions and trends, including the actions of competitors and consumer trends.

[1216] "Store data" is data related to a specific store, and includes store sales information, inventory status, customer demographics, and the like.

[1217] "Salesperson data" refers to data relating to individual salespersons, including sales performance, skill level, working hours, and the like.

[1218] "Preprocessing" refers to processes such as filling in missing values ​​in collected data, deleting duplicate data, and standardizing formats.

[1219] "Feature extraction" is the process of finding useful patterns and trends in the data being analyzed.

[1220] A "role play scenario" is a simulated sales situation scenario that a salesperson performs for learning and training purposes.

[1221] The "emotion engine" is an engine that analyzes the facial expressions and tone of voice of salespeople to recognize their emotions.

[1222] A "machine learning model" is an algorithm or system that learns from large amounts of data and performs pattern recognition and prediction.

[1223] "Feedback" refers to information such as evaluations, impressions, and areas for improvement collected from salespeople after the role-play is completed.

[1224] In this invention, a server, a terminal, and a user (salesperson) work together to realize a training system for salespersons. Specific embodiments for carrying out the invention are described below.

[1225] First, the server collects sales data, market data, store data, and salesperson data. This is done using APIs and database queries to obtain various data and store it in the database. It is also possible to obtain real-time data from external data sources.

[1226] Next, the server preprocesses the collected data, completing missing data, removing duplicate data, and standardizing the data format to prepare it for analysis. This preprocessing improves the accuracy of the analysis and enables more reliable feature extraction.

[1227] Based on the preprocessed data, the server performs data analysis and extracts characteristics of salespeople and stores. For example, it uses statistical methods and machine learning models to discover patterns such as which salespeople are effective on certain days or at certain times of the day, or which stores are effective in particular products.

[1228] After the features are extracted, the server generates role-play scenarios, leveraging neural networks and other generative AI models to generate scenarios optimized for each salesperson and store, including the skills and situations needed to improve the salesperson's performance.

[1229] The generated scenario is distributed from the server to the salesperson's device. For example, the scenario is sent to a device such as a smartphone or head-mounted display (HMD) via notification or file transfer. The salesperson then role-plays based on the distributed scenario.

[1230] At this time, the emotion engine installed in the device analyzes the salesperson's facial expressions and tone of voice to recognize their emotions. The emotion engine is responsible for analyzing in real time whether the salesperson is feeling stressed or relaxed. For example, if the server detects that the salesperson is feeling stressed, it can dynamically adjust the difficulty of the scenario based on that information.

[1231] After the role-play is completed, the user (salesperson) enters feedback into the terminal. The feedback includes the results of the role-play, impressions, and areas for improvement. This feedback is sent to the server and used to update the machine learning model. This makes it possible to provide even more accurate and effective scenarios the next time a scenario is generated.

[1232] For example, a scenario can be generated using the following prompt:

[1233] "Generate a scenario to improve closing skills for a female in her twenties who visits your store during lunchtime on a weekday. If the salesperson feels stressed, adjust the difficulty of the scenario to be less difficult."

[1234] Based on this prompt, the server can generate the optimal role-play scenario for specific conditions and skill sets.

[1235] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1236] Step 1: Data collection

[1237] The server collects sales data, market data, store data, and salesperson data using external APIs and database queries. It uses API endpoints and database connection information as input, retrieves the latest data from each data source, and stores it in the database. The output is a collection of the collected raw data.

[1238] Step 2: Data Preprocessing

[1239] The server preprocesses the collected raw data. It uses the collected raw data as input, complements missing data, removes duplicate data, and standardizes the format to make it suitable for analysis. The output is preprocessed clean data. Specifically, it finds defects in the data, complements them, and unifies inconsistent data formats.

[1240] Step 3: Data analysis

[1241] The server analyzes the preprocessed data and extracts features. It uses the preprocessed clean data as input. It uses statistical methods and machine learning models to analyze patterns of salespeople and stores. The output is feature data for each salesperson and store. Specifically, it finds patterns of salespeople who are strong on specific days of the week or at specific times of the day, or stores with high sales.

[1242] Step 4: Role-play scenario generation

[1243] The server generates role-play scenarios based on the extracted feature data. Utilizing neural networks and generative AI models, it designs scenarios optimized for each salesperson. Feature data and prompts are used as input. The output is the generated role-play scenario. Specifically, it automatically generates scenarios that strengthen the salesperson's weaknesses.

[1244] Step 5: Scenario distribution

[1245] The server distributes the generated scenario to the salesperson's terminal. It uses the generated role-play scenario as input. It sends the scenario to the terminal using notification or file transfer. The output is a notification or file that the scenario was distributed. Specifically, it sends a notification to the terminal so that the salesperson can check the received scenario.

[1246] Step 6: Sentiment Analysis

[1247] The device uses an emotion engine to analyze the salesperson's emotions during role-playing. Data on the salesperson's facial expressions and tone of voice are used as input. The emotion engine analyzes this data and recognizes emotional states such as stress or relaxation. The output is the recognized emotion data. Specifically, the device uses a camera and microphone to capture facial expressions and voice data in real time, which the emotion engine analyzes.

[1248] Step 7: Dynamically adjust the scenario

[1249] The server adjusts the difficulty and content of the scenario based on the emotional data sent from the terminal. Emotional data is used as input. If the salesperson is feeling stressed, the server adjusts the difficulty of the scenario, for example, by lowering it. The output is the adjusted scenario. Specifically, the server uses a difficulty adjustment algorithm to dynamically change the content of the scenario.

[1250] Step 8: Gather feedback

[1251] The terminal collects feedback from the salesperson after the role-play ends. The input is the feedback data entered by the user. The output is the collected feedback data. In concrete terms, the salesperson fills out a feedback form on the terminal, which is then sent to the server.

[1252] Step 9: Update the machine learning model

[1253] The server updates the machine learning model based on the collected feedback data. It uses the feedback data and past data as input. The updated model can provide more accurate scenarios the next time it generates a scenario. The output is an updated machine learning model. Specifically, it adds the new feedback data to the learning data and retrains the model.

[1254] 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.

[1255] 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.

[1256] 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.

[1257] 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.

[1258] 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.

[1259] 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.

[1260] 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).

[1261] 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.

[1262] 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."

[1263] 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.

[1264] 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).

[1265] 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.

[1266] 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.

[1267] 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.

[1268] 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.

[1269] 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.

[1270] 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.

[1271] 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.

[1272] 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.

[1273] 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.

[1274] 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.

[1275] The following is further disclosed regarding the above embodiment.

[1276] (Claim 1)

[1277] means for collecting sales data, market data, store data, and salesperson data;

[1278] means for pre-processing the collected data;

[1279] means for analyzing and extracting features from the preprocessed data;

[1280] means for generating a role-play scenario based on the extracted features;

[1281] means for distributing the generated scenario to a terminal of each salesperson;

[1282] a means for a salesperson to role-play based on the scenario and collect feedback;

[1283] means for updating a machine learning model based on the collected feedback.

[1284] (Claim 2)

[1285] 10. The system of claim 1, wherein the collected data includes sales data, market data, store data, and salesperson data.

[1286] (Claim 3)

[1287] 10. The system of claim 1, wherein the role-play scenario is customized based on salesperson characteristics and store characteristics.

[1288] "Example 1"

[1289] (Claim 1)

[1290] means for collecting sales data, market data, store data, and salesperson data;

[1291] means for pre-processing the collected data;

[1292] means for analyzing the pre-processed data using statistical techniques and machine learning algorithms to extract features;

[1293] a means for generating a role-play scenario using a neural network or the like based on the extracted features;

[1294] means for distributing the generated scenario to a terminal of each salesperson;

[1295] A means for salespeople to use the device to conduct role-plays and gather feedback;

[1296] means for analyzing the collected feedback and updating a machine learning model.

[1297] (Claim 2)

[1298] 10. The system of claim 1, wherein the collected data includes sales data, market data, store data, and salesperson data.

[1299] (Claim 3)

[1300] 10. The system of claim 1, wherein the role-play scenario is customized based on salesperson characteristics and store characteristics.

[1301] "Application Example 1"

[1302] (Claim 1)

[1303] means for collecting sales data, market data, store data, and salesperson data;

[1304] means for pre-processing the collected data;

[1305] means for analyzing and extracting features from the preprocessed data;

[1306] means for generating a role-play scenario based on the extracted features;

[1307] means for distributing the generated scenario to a terminal of each salesperson;

[1308] a means for a salesperson to role-play based on the scenario and collect feedback;

[1309] means for updating a machine learning model based on the collected feedback;

[1310] A means for generating scenarios for salespeople to improve their closing skills for specific products;

[1311] a means for automatically evaluating the content of responses in the role play using a voice recognition function of the terminal;

[1312] The system includes a means for utilizing feedback data to reflect it in the next scenario generation.

[1313] (Claim 2)

[1314] 10. The system of claim 1, wherein the collected data includes sales data, market data, store data, and salesperson data.

[1315] (Claim 3)

[1316] 10. The system of claim 1, wherein the role-play scenario is customized based on salesperson characteristics and store characteristics.

[1317] "Example 2: Combining Emotion Engines"

[1318] (Claim 1)

[1319] means for collecting sales data, market data, store data, and salesperson data;

[1320] means for pre-processing the collected data;

[1321] means for analyzing and extracting features from the preprocessed data;

[1322] means for generating role-play scenarios based on said extracted features using a neural network or other generative AI model;

[1323] means for distributing the generated scenario to a terminal of each salesperson;

[1324] A means for a salesperson to perform a role play based on the scenario and recognize emotions using an emotion engine of the terminal;

[1325] means for dynamically adjusting the difficulty and content of a role-play scenario based on said emotional information;

[1326] means for collecting feedback after the scenario is implemented;

[1327] means for updating a machine learning model based on the collected feedback.

[1328] (Claim 2)

[1329] 10. The system of claim 1, wherein the collected data includes sales data, market data, store data, and salesperson data.

[1330] (Claim 3)

[1331] 10. The system of claim 1, wherein the role-play scenario is customized based on salesperson characteristics and store characteristics.

[1332] "Application example 2 when combining emotion engines"

[1333] (Claim 1)

[1334] means for collecting sales data, market data, store data and salesperson data;

[1335] means for pre-processing the collected data;

[1336] means for analyzing and extracting features from the preprocessed data;

[1337] means for generating a role-play scenario based on the extracted features;

[1338] means for distributing the generated scenario to a terminal including means for analyzing emotions of a salesperson;

[1339] a means for having a salesperson perform a role play based on the scenario, adjusting the difficulty and content of the scenario based on the emotion analysis result, and collecting feedback;

[1340] The system further includes a means for updating the machine learning model based on the collected feedback and generating the next scenario.

[1341] (Claim 2)

[1342] 2. The system of claim 1, wherein the means for detecting the salesperson's emotions uses an emotion engine that analyzes facial expressions and tone of voice.

[1343] (Claim 3)

[1344] 10. The system of claim 1, wherein the role-play scenario is customized based on salesperson characteristics and store characteristics, and is dynamically adjusted based on sentiment analysis results. [Explanation of symbols]

[1345] 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. means for collecting sales data, market data, store data, and salesperson data; means for pre-processing the collected data; means for analyzing and extracting features from the preprocessed data; means for generating a role-play scenario based on the extracted features; means for distributing the generated scenario to a terminal of each salesperson; a means for a salesperson to role-play based on the scenario and collect feedback; means for updating a machine learning model based on the collected feedback.

2. 2. The system of claim 1, wherein the collected data includes sales data, market data, store data, and salesperson data.

3. The system of claim 1 , wherein the role-play scenario is customized based on salesperson characteristics and store characteristics.

Images