System

Abstract

A system is provided.SOLUTION: This system is provided with a user interface means for inputting a problem, a natural language processing means for analyzing the inputted problem, a means for generating a plurality of virtual knowledge agents having different thinking patterns, a means for starting a discussion by the virtual knowledge agents, and for collecting discussion results, and a means for outputting the discussion results to the user interface means.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] It is well-known that discussions can help avoid siloed thinking in problem-solving. However, depending on the circumstances of a company or team, people with similar ways of thinking may gather together, resulting in insufficient discussions. This makes it difficult to reach appropriate conclusions from diverse perspectives and find optimal solutions. The purpose of this invention is to solve this problem, deriving appropriate conclusions through discussions from diverse perspectives, thereby reducing discussion costs and improving work efficiency. [Means for solving the problem]

[0005] The present invention solves the aforementioned problems by a system comprising:

[0006] The system includes a user interface for inputting a problem, a natural language processing system for analyzing the input problem, a system for generating a plurality of virtual knowledge agents with different thought patterns, a system for initiating a discussion among the virtual knowledge agents, a system for aggregating the results of the discussion, and a system for outputting the results of the discussion to the user interface. This system enables appropriate conclusions to be reached through discussions from a variety of perspectives, reducing the cost of discussions and improving work efficiency.

[0007] Specifically, when a task is input, a text analysis of the task is performed using natural language processing means, and key keywords and intent are extracted. After that, multiple virtual knowledge agents are generated, and each agent is assigned a different thinking pattern (e.g., maximize sales, maximize profits, minimize costs). The virtual knowledge agents then begin a discussion based on the task, exchanging opinions from their respective perspectives. Finally, the results of the discussion are compiled and output to the user via a user interface means.

[0008] "User interface means" refers to an interface through which a user can input information to the system and check output from the system.

[0009] "Natural language processing means" refers to processing means that uses technology to analyze text entered by a user and extract keywords and intent.

[0010] A "virtual knowledge agent" is a virtual entity that has a specific thought pattern and is able to express opinions and discuss issues within the system.

[0011] "Thinking patterns" refer to specific perspectives or policies held by virtual knowledge agents, examples of which include sales maximization, profit maximization, and cost minimization.

[0012] "Means for initiating discussions and summarizing the results of discussions" refers to the processing means by which virtual knowledge agents exchange opinions about issues, summarize the results, and derive optimal conclusions.

[0013] "Input problem" refers to a specific problem or theme that a user inputs into the system to seek a solution.

[0014] "Key keywords and intent" refers to important elements necessary for solving a problem, extracted from the input problem text using natural language processing technology.

[0015] "Means for outputting the results of the discussion to the user interface means" refers to processing means for presenting the conclusions and proposals reached through the discussions between the virtual knowledge agents to the user through the user interface.

[0016] "System" refers to the entire combination of computer hardware and software that performs a series of operations such as inputting tasks, analyzing, discussing, and outputting results. [Brief explanation of the drawings]

[0017] [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

[0018] 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.

[0019] First, the terms used in the following description will be explained.

[0020] 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).

[0021] 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.

[0022] 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.

[0023] 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.

[0024] 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."

[0025] [First embodiment]

[0026] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0027] 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.

[0028] 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).

[0029] 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.

[0030] 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.

[0031] 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.

[0032] 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.

[0033] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0034] 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.

[0035] 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.

[0036] 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.

[0037] 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."

[0038] The present invention is a system in which a user inputs a problem through a terminal, and virtual knowledge agents with different thought patterns discuss the problem and derive an appropriate conclusion. The following describes in detail an embodiment of the present invention.

[0039] System Overview

[0040] 1. Enter your assignment

[0041] The user inputs the problem to be solved in text format using a terminal. For example, the user might input, "I want to determine the pricing strategy for new product A."

[0042] 2. Parsing the Input

[0043] The terminal receives the task input by the user and transmits it to the server.

[0044] The server analyzes the submitted text using natural language processing to extract key keywords and intent. Specifically, keywords such as "new product A," "pricing strategy," and "entering market" are extracted.

[0045] 3. Generation of virtual knowledge agents with diverse thinking patterns

[0046] The server generates virtual knowledge agents from an internal database, each with a different thought pattern.

[0047] Virtual Knowledge Agent 1: Thinking Patterns for Maximizing Sales

[0048] Virtual Knowledge Agent 2: Profit Maximization Thinking Pattern

[0049] Virtual Knowledge Agent 3: Cost Minimization Thinking Pattern

[0050] 4. Individual analysis of the issues

[0051] The server presents a task to each virtual knowledge agent and has them perform an analysis individually.

[0052] Virtual Knowledge Agent 1 considers the merits of a high-price strategy based on market data.

[0053] Virtual knowledge agent 2 considers the optimal profit margin based on revenue data and cost structure.

[0054] Virtual knowledge agent 3 analyzes strategies to minimize manufacturing and marketing costs.

[0055] 5. Initiating discussions and summarizing the results

[0056] The server initiates a discussion among the virtual knowledge agents on the topic, and each agent expresses their opinion from their own perspective and engages in the discussion.

[0057] The server aggregates these discussion results and derives the optimal conclusion. For example, the following discussion result is derived:

[0058] After considering the balance between whether to adopt a high-price strategy, a mid-price range, or a low-price strategy, it is determined that the mid-price range is most appropriate.

[0059] 6. Outputting the results

[0060] The server outputs the aggregated discussion results to the user's terminal through a user interface. As a specific example, the server may present to the user a conclusion such as setting the price of new product A in the mid-range to minimize manufacturing and marketing costs.

[0061] Specific examples

[0062] Consider the example of deciding on a pricing strategy for new product A. When a user types "I want to decide on a pricing strategy for new product A" into their terminal, the system proceeds according to the above procedure. Important keywords are extracted using natural language processing, and virtual knowledge agents discuss pricing strategies from their own perspectives. Ultimately, the server outputs the conclusion that "the price of new product A should be set in the mid-range, minimizing manufacturing and marketing costs," and this is displayed on the user's terminal.

[0063] This system allows appropriate conclusions to be reached through discussions from multiple perspectives, reducing the cost of discussions and improving work efficiency.

[0064] The processing flow will be explained below.

[0065] Step 1:

[0066] The user uses a terminal to input a problem in text format, such as "I want to decide on a pricing strategy for new product A."

[0067] Step 2:

[0068] The terminal receives the assignment text entered by the user and transmits this input data to the server.

[0069] Step 3:

[0070] To analyze the assignment text received by the server, natural language processing is used to perform text analysis. Specifically, key keywords and intents (e.g., "New Product A," "pricing strategy," "entering market," etc.) are extracted from the assignment text.

[0071] Step 4:

[0072] Based on the extracted keywords and intentions, the server generates virtual knowledge agents with different thought patterns from its internal database. Each agent is assigned the following thought patterns:

[0073] Virtual Knowledge Agent 1: Maximizing Sales

[0074] Virtual Knowledge Agent 2: Profit Maximization

[0075] Virtual Knowledge Agent 3: Cost Minimization

[0076] Step 5:

[0077] The server presents a task to each virtual knowledge agent and instructs them to perform an individual analysis. Each agent performs an individual analysis as follows:

[0078] Virtual Knowledge Agent 1 considers the merits of setting a higher price based on market data.

[0079] Virtual Knowledge Agent 2 considers the optimal profit margin based on revenue data and cost structure.

[0080] Virtual knowledge agent 3 considers strategies to minimize manufacturing and marketing costs.

[0081] Step 6:

[0082] The server starts a discussion among the virtual knowledge agents. This allows each agent to express their opinion from their own perspective. The specific flow of the exchange of opinions is as follows:

[0083] Virtual Knowledge Agent 1: "New Product A should be sold at a high price and targeted at the premium market to maximize sales."

[0084] Virtual Knowledge Agent 2: "To maximize profits, we should set the price at a reasonable level relative to our competitors' products."

[0085] Virtual Knowledge Agent 3: "We should adopt a low-price strategy when entering the market to keep our manufacturing and marketing costs low."

[0086] Step 7:

[0087] The server aggregates the results of the discussions among the virtual knowledge agents and derives an overall conclusion, such as "It is optimal to set the price of new product A in the mid-range and minimize manufacturing and marketing costs."

[0088] Step 8:

[0089] The server transmits the aggregated discussion results to the terminal through the user interface means, allowing the user to check the final conclusions and proposals.

[0090] Step 9:

[0091] The terminal displays the final conclusion sent from the server to the user. The user confirms the proposal, "Price new product A at a mid-range price to minimize manufacturing and marketing costs," and takes the proposal into consideration.

[0092] Through the above steps, the system of the present invention derives appropriate conclusions through discussions from a variety of perspectives, thereby improving business efficiency and reducing discussion costs.

[0093] Example 1

[0094] 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."

[0095] Conventional systems lack the ability to discuss the issues users face from multiple perspectives, making it difficult to reach optimal conclusions. In particular, the lack of virtual knowledge agents with different thinking patterns means that analysis is limited to a single perspective, which can lead to overlooking various approaches that may be useful in resolving the issue.

[0096] 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.

[0097] In this invention, the server includes a user interface means for inputting a problem, a natural language processing means for analyzing the input problem, a means for generating a plurality of virtual knowledge agents with different thought patterns, a means for presenting a problem to the virtual knowledge agents and having them analyze it individually, a means for initiating a discussion among the virtual knowledge agents and aggregating the results of the discussion, and a means for outputting the results of the discussion to the user interface means, thereby enabling users to reach an optimal conclusion through discussion from multiple perspectives.

[0098] A "problem" is a specific problem or theme that a user seeks to solve.

[0099] "User interface means" refers to the input / output interface used by the user to input tasks and display results.

[0100] "Natural language processing means" refers to technical means for analyzing natural language used by humans and extracting key keywords and intent.

[0101] A "virtual knowledge agent" is an artificial intelligence virtual agent that has a different thought pattern and can perform individual analysis of a problem.

[0102] A "discussion method" is a means by which multiple virtual knowledge agents exchange opinions from their own perspectives and carry out a process to arrive at an optimal conclusion.

[0103] The "aggregation means" is a technical means for collecting and integrating the discussion results of each virtual knowledge agent to derive a final conclusion.

[0104] "Output means" refers to the means used to present the results of the discussion to the user.

[0105] The present invention is a system in which a user inputs a problem through a terminal, and virtual knowledge agents with different thought patterns discuss the problem and derive an appropriate conclusion. An embodiment of this system will be described in detail below.

[0106] System Overview

[0107] The system includes the following main elements:

[0108] 1. User Interface Methods

[0109] The user inputs the task in text format. The terminal receives this input and provides an interface for sending it to the server. The user interface is implemented as a web browser or a dedicated application.

[0110] 2. Natural Language Processing Methods

[0111] The server receives the assignment text sent by the user and analyzes it using natural language processing techniques (e.g., Python's NLTK library). This analysis extracts key keywords and intent. For example, keywords such as "New Product A," "Pricing Strategy," and "Market Entry" are extracted.

[0112] 3. Creation of Virtual Knowledge Agents

[0113] The server generates virtual knowledge agents with different thought patterns from an internal database (e.g., MySQL), including different perspectives such as sales maximization, profit maximization, and cost minimization.

[0114] Virtual Knowledge Agent 1: Thinking Patterns for Maximizing Sales

[0115] Virtual Knowledge Agent 2: Profit Maximization Thinking Pattern

[0116] Virtual Knowledge Agent 3: Cost Minimization Thinking Pattern

[0117] 4. Individual analysis of the issues

[0118] The server presents each virtual knowledge agent with a task entered by the user and has them perform an individual analysis. Each agent performs the analysis based on relevant data sets (e.g., market research data, revenue data, manufacturing cost data).

[0119] 5. Initiating discussions and summarizing the results

[0120] The server initiates a discussion based on the analysis results of virtual knowledge agents with different thought patterns, and aggregates the results. This discussion is conducted via an API, sharing the data and perspectives of each agent to arrive at the optimal conclusion.

[0121] 6. Outputting the results

[0122] The server outputs the results of the discussion to the user interface and displays them on the user's device. For example, a conclusion such as "Price new product A should be set in the mid-range, minimizing manufacturing and marketing costs" may be presented.

[0123] Specific examples

[0124] Consider a scenario in which a pricing strategy for new product A is to be decided. A user inputs "I would like to decide on a pricing strategy for new product A" into their device. This problem is sent to a server, where important keywords are extracted using natural language processing technology. The server then generates virtual knowledge agents, each of which analyzes the problem from a different perspective. As a result, the agents debate each other, and the final conclusion is displayed on the user's device.

[0125] Prompt Sentence Examples

[0126] Below are some example prompts that can be input to a generative AI model:

[0127] "I'd like to decide on a pricing strategy for new product A. What are the advantages and disadvantages of a high-price strategy, a mid-price strategy, and a low-price strategy?"

[0128] "Please tell us the results of the discussions among the virtual knowledge agents regarding the marketing strategy for new product B."

[0129] The above prompts allow the user to understand how to operate the system and the output results through specific scenarios.

[0130] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0131] Step 1:

[0132] The user inputs the problem to be solved in text format into the terminal. By typing "I want to decide on the pricing strategy for new product A" in the input field and pressing the send button, the input problem is sent from the terminal to the server.

[0133] Step 2:

[0134] The terminal receives the assignment entered by the user and sends it to the server via an HTTP request. The server prepares the received assignment text data for analysis. The input data is in text format, and the server queues this data for analysis.

[0135] Step 3:

[0136] The server performs natural language processing on the received assignment text using Python's NLTK library. Here, the text is subjected to morphological analysis to extract key keywords and intent. This analysis extracts keywords such as "new product A," "pricing strategy," and "market entry." The output is a list of the extracted keywords.

[0137] Step 4:

[0138] The server generates virtual knowledge agents with different thinking patterns from an internal database (MySQL) based on the extracted keywords. Each agent is set with a thinking pattern of maximizing sales, maximizing profits, or minimizing costs. For example, virtual knowledge agent 1 has a thinking pattern of maximizing sales, while virtual knowledge agent 2 has a thinking pattern of maximizing profits. The output is instances of the multiple virtual knowledge agents generated.

[0139] Step 5:

[0140] The server presents the tasks entered by the user to the generated virtual knowledge agents and has them perform individual analyses. Each agent references related data sets (e.g., market research data or revenue data) and performs analysis based on its own thought patterns. For example, virtual knowledge agent 1 considers a high-price strategy based on market data. The output is a report of the analysis results by each agent.

[0141] Step 6:

[0142] The server initiates a discussion based on the analysis results of the virtual knowledge agents. In this discussion, the virtual knowledge agents exchange opinions via API and consider the issue from multiple perspectives. Each agent shares their analysis results and perspectives, and a discussion is held to arrive at the optimal conclusion. The output is the conclusion agreed upon through the discussion.

[0143] Step 7:

[0144] The server aggregates the results of the discussions and derives a final conclusion. For example, the conclusion may be, "The price of new product A should be set in the mid-range, minimizing manufacturing and marketing costs." This conclusion is prepared as text data.

[0145] Step 8:

[0146] The server sends the aggregated discussion results to the user interface, which displays them on the user's terminal. The user interface presents the results to the user in an easy-to-read format, for example, in the form of a dashboard or report. The output is the final conclusion displayed on the user's terminal.

[0147] By following the above steps, users can reach optimal conclusions through discussions from multiple perspectives.

[0148] (Application example 1)

[0149] 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."

[0150] Conventional content delivery systems have the problem of being unable to quickly and accurately recommend content that meets the detailed requirements of users. In particular, recommendations that take into account users' interests and viewing history require analysis and discussion from multiple perspectives. Therefore, an efficient and comprehensive method for providing optimal content to users is required.

[0151] 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.

[0152] In this invention, the server includes a user interface means for inputting a problem, a natural language processing means for analyzing the input problem, a means for generating a plurality of virtual knowledge agents with different thought patterns, a means for initiating a discussion among the virtual knowledge agents and aggregating the results of the discussion, a means for outputting the results of the discussion to the user interface means, a means for recommending content based on the problem input by the user, and a means for the virtual knowledge agents to analyze and evaluate the content from their own different perspectives. This makes it possible to recommend optimal content to users through comprehensive discussion and analysis from different perspectives.

[0153] "Issues" refer to problems or requests that users want to solve.

[0154] "User interface means" refers to an interface through which a user inputs tasks and receives results.

[0155] "Natural language processing means" refers to technical means that analyze input text and extract key keywords and intent.

[0156] A "virtual knowledge agent" refers to a virtual agent that has a specific thought pattern and analyzes and discusses issues.

[0157] "Discussion results" refers to the output obtained as the conclusion of a discussion conducted by multiple virtual knowledge agents.

[0158] "Emotion analysis" refers to the perspective of analyzing a user's emotions and psychological state.

[0159] "Trend following" refers to a perspective based on current trends and popularity.

[0160] "Data-driven" refers to a perspective based on a user's past data and viewing history.

[0161] "Means for recommending content" refers to means for recommending optimal content based on user input.

[0162] MODE FOR CARRYING OUT THE INVENTION

[0163] System Program

[0164] The system program for realizing this invention is a system in which virtual knowledge agents with different thought patterns recommend content based on the user's issues. This system is composed of a user interface means, natural language processing means, virtual knowledge agent generation means, discussion result aggregation means, and result output means.

[0165] Natural language explanation of the process

[0166] Hardware and Software Configuration

[0167] In this system, users input tasks using a smartphone or a head-mounted display (HMD). On the server side, we use the Hugging Face transformers library to perform natural language processing. The creation of virtual knowledge agents and the aggregation of discussions are implemented using Python code.

[0168] Data processing and calculation

[0169] 1. The user inputs a task via a smartphone or HMD. For example, the user inputs the text "I want to watch a relaxing movie."

[0170] 2. The device sends the input text to the server, which uses Hugging Face's transformers library for natural language processing to extract key keywords and intent.

[0171] 3. The server generates multiple virtual knowledge agents with different thinking patterns, including sentiment analysis, trend following, and data-driven perspectives.

[0172] 4. The virtual knowledge agents perform individual analyses of the input tasks from their own perspectives. For example, a sentiment analysis agent might recommend relaxing movies, while a trend-following agent might recommend the latest popular movies.

[0173] 5. The server aggregates the discussion results of each agent and generates an optimal content recommendation list.

[0174] 6. Finally, the results are output to the user via a smartphone or HMD, and the user can receive a list of recommended content.

[0175] Specific examples

[0176] As a concrete example, consider the case where a user enters "I want to watch a relaxing movie." In this case, the system will process it through the following steps:

[0177] User: "I want to watch a relaxing movie."

[0178] Emotion Agent: "This movie is relaxing."

[0179] Trend Agent: "Here are your latest favorite movies."

[0180] Data-Driven Agent: "I found some relaxing movies based on your viewing history."

[0181] In this way, optimal content is recommended to the user after comprehensive discussion from multiple perspectives.

[0182] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0183] Step 1:

[0184] The user inputs a task using a smartphone or a head-mounted display (HMD). For example, the user inputs text such as "I want to watch a relaxing movie." This input is sent to the system as a task.

[0185] Input: Assignment text "I want to watch a relaxing movie"

[0186] Output: Send the assignment text to the server

[0187] Step 2:

[0188] The terminal transmits the task text entered by the user to the server, which receives the text for natural language processing.

[0189] Input: Assignment text

[0190] Output: Assignment text received on the server

[0191] Step 3:

[0192] The server uses Hugging Face's transformers library to perform natural language analysis of the challenge text and extract key keywords and intent, such as "relax" and "movie."

[0193] Input: Assignment text

[0194] Output: Keywords "relax", "movie"

[0195] Step 4:

[0196] The server generates multiple virtual knowledge agents with different thinking patterns, such as sentiment analysis, trend following, and data-driven thinking.

[0197] Input:keyword

[0198] Output: Sentiment analysis agents, trend-following agents, data-driven agents

[0199] Step 5:

[0200] Each virtual knowledge agent performs a separate analysis of the input task from its own perspective. For example, a sentiment analysis agent recommends movies with a relaxing effect, a trend-following agent recommends the latest popular movies, and a data-driven agent recommends relaxing movies based on the user's viewing history.

[0201] Input:keyword

[0202] Output: Recommendation results of sentiment analysis agents, recommendation results of trend follower agents, recommendation results of data-driven agents

[0203] Step 6:

[0204] The server aggregates the discussion results of each agent and creates an optimal content recommendation list, which incorporates the recommendations of each agent.

[0205] Input: Recommendation results of each agent

[0206] Output: Best content recommendation list

[0207] Step 7:

[0208] Finally, the server outputs a list of optimal content recommendations to the user via their smartphone or HMD, allowing the user to select the most appropriate content from the list and watch it.

[0209] Input: Best content recommendation list

[0210] Output: Display a list of content recommendations to the user

[0211] 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.

[0212] The present invention is a system that derives more appropriate discussion results by combining an emotion engine that recognizes the emotional state of the user. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The following describes in detail the embodiments of the present invention.

[0213] System Overview

[0214] 1. Enter your assignment

[0215] The user uses a terminal to input a problem in text format, such as "I want to decide on a pricing strategy for new product A." At the same time, the user's emotional state is recognized.

[0216] 2. Parsing the Input

[0217] The terminal receives the task text and emotion data input by the user and transmits them to the server.

[0218] The server analyzes the submitted task text using natural language processing to extract key keywords and intent. Specifically, keywords such as "new product A," "pricing strategy," and "entering market" are extracted. Emotion data provided by the emotion engine is also taken into consideration.

[0219] 3. Generation of virtual knowledge agents with diverse thinking patterns

[0220] The server generates virtual knowledge agents from its internal database. Each agent is assigned a different thought pattern:

[0221] Virtual Knowledge Agent 1: Maximizing Sales

[0222] Virtual Knowledge Agent 2: Profit Maximization

[0223] Virtual Knowledge Agent 3: Cost Minimization

[0224] 4. Individual analysis of the issues

[0225] The server presents a task to each virtual knowledge agent and instructs them to perform an individual analysis. Each agent performs an individual analysis as follows:

[0226] Virtual Knowledge Agent 1 considers the merits of setting a higher price based on market data.

[0227] Virtual Knowledge Agent 2 considers the optimal profit margin based on revenue data and cost structure.

[0228] Virtual knowledge agent 3 considers strategies to minimize manufacturing and marketing costs.

[0229] 5. Initiating discussions and summarizing the results

[0230] The server initiates a discussion among the virtual knowledge agents, allowing each agent to express their own opinion from their own perspective. During the discussion, the emotion engine continuously monitors the user's emotional state and provides emotional data in real time. This allows the virtual knowledge agents to adaptively adjust the content of the discussion taking into account the emotional data.

[0231] The specific flow of the exchange of opinions is as follows:

[0232] Virtual Knowledge Agent 1: "New Product A should be sold at a high price and targeted at the premium market to maximize sales."

[0233] Virtual Knowledge Agent 2: "To maximize profits, we should set the price at a reasonable level relative to our competitors' products."

[0234] Virtual Knowledge Agent 3: "We should adopt a low-price strategy when entering the market to keep our manufacturing and marketing costs low."

[0235] 6. Aggregating discussion results and outputting them to the user

[0236] The server aggregates the results of the discussions and derives a comprehensive conclusion that takes into account sentiment data. For example, the following conclusion may be derived: "It is optimal to set the price of new product A in the mid-range and minimize manufacturing and marketing costs."

[0237] The server transmits the aggregated discussion results to the terminal through the user interface means, allowing the user to check the final conclusions and proposals.

[0238] 7. Displaying the results

[0239] The terminal displays the final conclusion sent from the server to the user. The user confirms the proposal, "Price new product A at a mid-range price to minimize manufacturing and marketing costs," and takes the proposal into consideration.

[0240] Specific examples

[0241] For example, if a user inputs a task to determine a pricing strategy for new product A, the system will proceed according to the procedure described above. In particular, if the user indicates emotions such as "tension" or "anxiety" when inputting, the emotion engine will detect this and the virtual knowledge agents will engage in discussions that will make the user feel more at ease, resulting in a conclusion that is easy for the user to accept. As a result, the optimal conclusion is provided to the user: "The price of new product A should be set in the middle range, while minimizing manufacturing and marketing costs."

[0242] This system uses an emotion engine to conduct discussions that take into account the user's emotional state, making it possible to derive more appropriate and practical conclusions.

[0243] The processing flow will be explained below.

[0244] Step 1:

[0245] The user uses the device to input a problem in text format, such as "I want to decide on a pricing strategy for new product A." At the same time, the user's emotional state is recognized from their facial expressions and voice.

[0246] Step 2:

[0247] The terminal receives the task text entered by the user and the emotion data analyzed by the emotion engine, and transmits them to the server.

[0248] Step 3:

[0249] To analyze the assignment text received by the server, natural language processing is used to perform text analysis. Specifically, key keywords and intent (e.g., "New Product A," "pricing strategy," "entering market," etc.) are extracted from the assignment text. At the same time, emotional data provided by the emotion engine is also taken into account in the analysis.

[0250] Step 4:

[0251] Based on the analysis results, the server generates virtual knowledge agents with different thought patterns from its internal database. Each agent is assigned the following thought patterns:

[0252] Virtual Knowledge Agent 1: Maximizing Sales

[0253] Virtual Knowledge Agent 2: Profit Maximization

[0254] Virtual Knowledge Agent 3: Cost Minimization

[0255] Step 5:

[0256] The server presents each virtual knowledge agent with a task and instructs them to perform an individual analysis. The specific analysis content is as follows:

[0257] Virtual Knowledge Agent 1 considers the merits of setting a higher price based on market data.

[0258] Virtual Knowledge Agent 2 considers the optimal profit margin based on revenue data and cost structure.

[0259] Virtual knowledge agent 3 considers strategies to minimize manufacturing and marketing costs.

[0260] Step 6:

[0261] The server initiates a discussion among the virtual knowledge agents. Each agent expresses their opinion from their own perspective as follows:

[0262] Virtual Knowledge Agent 1: "New Product A should be sold at a high price and targeted at the premium market to maximize sales."

[0263] Virtual Knowledge Agent 2: "To maximize profits, we should set the price at a reasonable level relative to our competitors' products."

[0264] Virtual Knowledge Agent 3: "We should adopt a low-price strategy when entering the market to keep our manufacturing and marketing costs low."

[0265] Step 7:

[0266] The server instructs the virtual knowledge agents to adaptively adjust the content of the discussion based on real-time emotional data provided by the emotion engine. For example, if the user is in a "tense" state, the agent will adjust the content of the discussion to make the user feel more at ease.

[0267] Step 8:

[0268] The server aggregates the results of the discussions and derives a comprehensive conclusion that takes into account sentiment data. For example, the following conclusion may be reached: "It is optimal to set the price of new product A in the mid-range and minimize manufacturing and marketing costs."

[0269] Step 9:

[0270] The server sends this conclusion to the terminal through the user interface means, allowing the user to check the final discussion results and conclusions.

[0271] Step 10:

[0272] The terminal displays the final conclusion sent from the server to the user, who confirms the proposal that "the price of new product A should be set in the mid-range to minimize manufacturing and marketing costs."

[0273] Through the detailed steps described above, the system of the present invention conducts discussions that take into account the user's emotional state, thereby improving work efficiency and deriving appropriate conclusions.

[0274] Example 2

[0275] 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."

[0276] Conventional systems were unable to consider the user's emotional state when deriving solutions to problems submitted by the user. As a result, even if practical and specific proposals were made, it was sometimes difficult to reach a conclusion that was easy for the user to accept. It was also difficult to derive a comprehensive conclusion through discussions from various perspectives. This has led to the need for a system that can provide the appropriate problem solutions that users desire.

[0277] 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.

[0278] In this invention, the server includes a user interface for inputting a task, a natural language processing unit for analyzing the input task and emotional data, a unit for generating a plurality of virtual knowledge agents with different thought patterns, a unit for initiating a discussion among the virtual knowledge agents, a unit for monitoring the emotional state of the user in real time during the discussion, and a unit for aggregating the results of the discussion, and a unit for outputting the results of the discussion to the user interface. This enables discussions that take the emotional state of the user into consideration, leading to more convincing conclusions. It is also possible to provide comprehensive proposals from the perspectives of a variety of virtual knowledge agents.

[0279] A "problem" is a specific problem or requirement that a user wants solved.

[0280] The "user interface means" is an interface through which a user inputs a task into the system.

[0281] "Natural language processing means" is a means for analyzing input text data and emotional data and extracting key keywords and intentions.

[0282] "Emotional data" is data that indicates the user's current emotional state and is based on information collected from facial expressions, voice, etc.

[0283] "Virtual Knowledge Agents" are virtual knowledge-based agents with different thinking patterns, each generated to analyze a problem according to a specific goal or policy.

[0284] "Discussion" is a process in which multiple virtual knowledge agents exchange opinions based on their own perspectives and seek optimal solutions.

[0285] "Discussion results" are the final conclusions or proposals derived from discussions among virtual knowledge agents.

[0286] The "means for outputting to the user interface means" is a means for allowing the user to visually confirm the results of the discussion.

[0287] "Real-time monitoring" refers to continuously observing the user's emotional state and instantly acquiring that data.

[0288] System Overview

[0289] The present invention is a system that derives more appropriate discussion results by combining an emotion engine that recognizes the emotional state of the user. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The following describes in detail the preferred embodiments of the present invention.

[0290] Hardware and Software Configuration

[0291] 1. Server

[0292] The server analyzes the input data using natural language processing tools (e.g., spaCy or NLTK) and generates a virtual knowledge agent using a generative AI model (e.g., GPT-4 or BERT) to facilitate the discussion.

[0293] The server is equipped with an emotion engine and is designed to monitor the user's emotional state in real time.

[0294] 2. Terminal

[0295] The terminal provides an interface for users to input tasks, and is equipped with features such as a text area, voice input, and a face recognition sensor.

[0296] The terminal is equipped with a communication module for transmitting the user's input data and emotion data to the server.

[0297] 3. Users

[0298] The user uses a terminal to input the issues and receive the review results.

[0299] The emotion engine recognizes the user's emotional state (e.g., "tension" or "anxiety") in real time.

[0300] Specific program behavior

[0301] 1. Task input and emotion recognition

[0302] The user uses the device to input a problem in text format, such as "I want to decide on a pricing strategy for new product A." At this time, sensors and cameras built into the device collect data on the user's emotions and analyze it in real time.

[0303] 2. Sending input data

[0304] The device sends the input text and emotion data to the server. The text data is packaged in JSON format, and the emotion data is encoded in a proprietary format.

[0305] 3. Analysis of input data

[0306] The server uses natural language processing tools to analyze the received text data and extract key keywords and intent. For example, keywords such as "new product A," "pricing strategy," and "entering market" are extracted. An emotion engine also interprets the emotional data and identifies states such as "tension" and "anxiety."

[0307] 4. Creation of Virtual Knowledge Agents

[0308] The server generates multiple virtual knowledge agents with different thinking patterns based on its internal database. For example, virtual knowledge agent 1 is set to maximize sales, virtual knowledge agent 2 to maximize profits, and virtual knowledge agent 3 to minimize costs.

[0309] Use generative AI models (e.g., GPT-4 or BERT) to build each agent's thought patterns.

[0310] 5. Performing individual analyses

[0311] The server requests each virtual knowledge agent to analyze a specific problem. For example, virtual knowledge agent 1 analyzes market data and considers the merits of setting high prices. Virtual knowledge agent 2 analyzes revenue data and cost structures to find the optimal profit margin. Virtual knowledge agent 3 analyzes manufacturing and marketing costs and proposes ways to minimize them.

[0312] 6. Initiating and Conducting a Discussion

[0313] The server initiates a discussion among the virtual knowledge agents, and each agent takes turns expressing their opinion. Virtual knowledge agent 1 states, "New product A should be sold at a high price and target the premium market," to which virtual knowledge agent 2 counters, "It should be priced at an appropriate level compared to competing products." During this time, the emotion engine continues to monitor the user's emotional state and provides data in real time.

[0314] 7. Summary of discussion results

[0315] The server aggregates the opinions of each agent and draws an overall conclusion taking into account the sentiment data. For example, it may conclude that "it is optimal to set the price of new product A in the mid-range and minimize manufacturing and marketing costs."

[0316] 8. Sending and displaying results to the user

[0317] The server sends the analysis results and conclusions of the discussion to the terminal via a user interface. The data is structured in JSON format and is displayed visually on the terminal.

[0318] The terminal displays the conclusion "The price of new product A should be set in the middle price range to minimize manufacturing and marketing costs" to the user, allowing the user to confirm the content.

[0319] Examples of concrete examples and prompts

[0320] For example, when a user inputs a task to determine a pricing strategy for new product A, the system proceeds according to the above procedure. In particular, if the user indicates emotions such as "tension" or "anxiety" when inputting, the emotion engine detects this, and the virtual knowledge agents engage in discussions that make the user feel more at ease, resulting in a conclusion that is easy for the user to accept. Ultimately, the optimal conclusion is provided to the user: "The price of new product A should be set in the middle range, minimizing manufacturing and marketing costs."

[0321] Example prompt sentence:

[0322] "Please suggest the optimal strategy to price New Product A in the mid-range and minimize manufacturing and marketing costs. Users are expressing uncertainty, so please provide a reassuring explanation."

[0323] This system uses an emotion engine to conduct discussions that take into account the user's emotional state, making it possible to derive more appropriate and practical conclusions.

[0324] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0325] Step 1:

[0326] User task input and emotion recognition

[0327] The user inputs a specific task (e.g., determining the pricing strategy for new product A) in text format through the terminal interface.

[0328] The device uses built-in sensors and cameras to collect emotional data in real time from the user's facial expressions, voice, etc.

[0329] Input: Task text, emotion data (facial expressions, voice, etc.)

[0330] Output: Prepare the combined task text and sentiment data as a data package for transmission.

[0331] How it works: The device receives user input, converts it into text format, and collects emotional data using sensors and compiles it into a single package.

[0332] Step 2:

[0333] Terminal sends input data

[0334] The device sends the combined task text and emotion data to the server, where the data is encoded in JSON or a dedicated format via a communication module.

[0335] Input: A data package containing the task text and sentiment data

[0336] Output: Data package sent to the server

[0337] How it works: The device sends data and the server receives it. Data transmission is designed to be highly reliable and with low latency.

[0338] Step 3:

[0339] Input analysis and emotional data acquisition by the server

[0340] The server parses the received data package. Natural language processing tools (e.g., spaCy or NLTK) are used to extract key keywords and intent from the assignment text. An emotion engine analyzes the emotion data and identifies the user's current emotional state (e.g., "nervous," "anxious," etc.).

[0341] Input: Issue text and sentiment data in JSON format

[0342] Output: Analysis results including extracted keywords and emotional states

[0343] How it works: Natural language processing tools perform text analysis, and the emotion engine processes emotion data in real time. For example, keywords such as "new product A" and "pricing strategy" are extracted, and the user's emotional state is identified as "anxiety."

[0344] Step 4:

[0345] Server-based virtual knowledge agent generation

[0346] The server generates multiple virtual knowledge agents with different thinking patterns (e.g., sales maximization, profit maximization, cost minimization) based on an internal database. Each agent is built using a generative AI model (e.g., GPT-4, BERT).

[0347] Input: Analysis results including extracted keywords and emotional states

[0348] Output: A list of virtual knowledge agents

[0349] Specific operation: Based on the analysis results, the server sets the thinking pattern of each agent and constructs the agent using an appropriate AI model.

[0350] Step 5:

[0351] Individual analysis performed by the server

[0352] The server requests each virtual knowledge agent to analyze a specific problem. For example, virtual knowledge agent 1 analyzes market data and considers the merits of setting a high price.

[0353] Input: List of virtual knowledge agents and analysis results

[0354] Output: The results of the analysis by each agent

[0355] Specific operation: Each agent retrieves the necessary information from the database and analyzes it based on its own thought pattern. For example, Agent 1 suggests, "New Product A should be sold at a high price and target the premium market."

[0356] Step 6:

[0357] Starting and running discussions by the server

[0358] The server initiates discussions among the virtual knowledge agents, and each agent expresses their opinion. The emotion engine monitors the user's emotional state in real time and provides the data.

[0359] Input: Analysis results and sentiment data for each agent

[0360] Output: Discussion progress and each agent's opinion

[0361] Specific operation: Agent 1 proposes a high price strategy, Agent 2 counters that the price should be set at an appropriate range, and Agent 3 emphasizes cost minimization. The emotion engine detects the user's "anxiety," and the agents adjust the content of the discussion based on that data.

[0362] Step 7:

[0363] Aggregation of discussion results by the server

[0364] The server aggregates the results of the discussions and derives an overall conclusion taking into account the emotional data.

[0365] Input: Discussion progress and sentiment data

[0366] Output: Overall conclusion

[0367] Specific operation: The server organizes the opinions of each agent and derives a conclusion, for example, that "it is optimal to set the price of new product A in the mid-range and minimize manufacturing and marketing costs."

[0368] Step 8:

[0369] The server sends the results to the user

[0370] The server sends the analysis results and conclusions of the discussion to the terminal via the user interface.

[0371] Input: Overall conclusion

[0372] Output: Conclusion data sent to the user terminal

[0373] Specific behavior: Data structured in JSON format is sent to the device.

[0374] Step 9:

[0375] Displaying results on a terminal

[0376] The terminal visually displays the results sent from the server to the user.

[0377] Input: Conclusion data sent from the server

[0378] Output: The final conclusion that is displayed to the user

[0379] Specific operation: The device displays the conclusion to the user, "New Product A should be priced in the mid-range to minimize manufacturing and marketing costs," and allows the user to confirm the conclusion.

[0380] (Application example 2)

[0381] 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."

[0382] When deciding on new product displays and pricing strategies in physical stores, employees and store managers face a variety of challenges, and there is a need to find more appropriate and effective solutions. However, decisions based on employees' individual experience and judgment are often influenced by emotions and intuition, making it difficult to derive optimal results. Furthermore, because it is not possible to take into account employees' emotional states, such as tension and anxiety, there is a lack of sufficient information to select the best strategy.

[0383] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0384] In this invention, the server includes a user interface means for inputting a problem, a natural language processing means for analyzing the input problem, a means for generating a plurality of virtual knowledge agents with different thought patterns, a means for initiating a discussion among the virtual knowledge agents and aggregating the discussion results, a means for outputting the discussion results to the user interface means, an emotion recognition means for recognizing the emotional state of the user, and a means for adaptively adjusting the discussion of the virtual knowledge agents based on the emotional state. This makes it possible to derive appropriate discussion results by the virtual knowledge agents while taking the emotional state of the user into consideration, and to provide optimal solutions for new product displays and pricing strategies in physical stores.

[0385] The "user interface means for inputting issues" refers to an interface that allows a user to input issues or requests that they wish to solve, and can be operated on the screen of a terminal.

[0386] The "natural language processing means for analyzing the input problem" is a means for analyzing the text input by the user and extracting the intention and main keywords.

[0387] "Multiple virtual knowledge agents with different thinking patterns" are virtual agents that analyze issues and offer opinions based on specific strategies or perspectives, and have different thinking patterns such as sales maximization, profit maximization, or cost minimization.

[0388] The "means for initiating a discussion among the virtual knowledge agents and summarizing the results of the discussion" is a means for a plurality of virtual knowledge agents to hold a discussion from their respective viewpoints and summarizing the results.

[0389] The "means for outputting the discussion results to the user interface means" refers to a means for displaying the summarized discussion results to the user.

[0390] The "emotion recognition means for recognizing the user's emotional state" is a means for recognizing the user's emotion from their facial expression or voice, and is used to understand the mental state of the user while they are inputting.

[0391] The "means for adaptively adjusting the discussion of the virtual knowledge agents based on the emotional state" refers to a means for adaptively adjusting the content and progress of the discussion of the virtual knowledge agents in consideration of the emotional state of the user.

[0392] To implement this invention, it is necessary to build a system that allows users to decide sales strategies for physical stores. This system includes the following components:

[0393] User Interface Means

[0394] The interface for users to input their tasks is a device such as a tablet or smartphone. This device provides a screen on which users can enter text about specific tasks, such as how to display a new product or a pricing strategy.

[0395] Natural language processing tools

[0396] To analyze the assignment text sent from the device, the server uses a natural language processing library (e.g., NaiveBayesClassifier), which allows it to extract key keywords and the intent of the assignment.

[0397] Virtual Knowledge Agent

[0398] The server generates multiple virtual knowledge agents with different thought patterns. For example, agents with perspectives such as maximizing sales, maximizing profits, and minimizing costs are generated. Each agent proposes the optimal strategy from its own perspective based on data analysis.

[0399] emotion recognition means

[0400] The device is equipped with a camera and uses an emotion recognition engine (e.g., EmotionRecognizer) to recognize the user's facial expressions and voice in real time, which makes it possible to understand the user's emotional state (tension, anxiety, etc.) while they are inputting.

[0401] Discussion and summary of results

[0402] The server instructs the virtual knowledge agents to hold discussions while taking into account the user's emotional state. The agents exchange opinions from different perspectives and ultimately derive an optimal conclusion. This result is displayed to the user through a user interface.

[0403] Specific examples

[0404] For example, if a user types "how to display new products" into the tablet, the system will function as follows:

[0405] 1. Task input: The user inputs "how to display a new product" on the tablet.

[0406] 2. Emotion Recognition: The emotion of "tension" is recognized through the camera while the user is typing.

[0407] 3. Natural Language Processing: The server analyzes the input text and extracts key keywords and intent.

[0408] 4. Virtual knowledge agent generation: Virtual knowledge agents with different thinking patterns are generated.

[0409] 5. Discussion and summary of results: The agents hold a discussion to reassure the nervous user and propose a strategy for displaying the product at eye level in the mid-price range.

[0410] 6. Displaying the results: The final proposal is displayed on the tablet, providing the user with specific advice such as "price the new product in the middle of the market and display it at eye level."

[0411] Prompt Sentence Examples

[0412] Example: Prompt text that the user types into the tablet

[0413] How to display new products

[0414] As described above, this invention recognizes the emotional state of the user and responds in real time, and virtual knowledge agents adaptively discuss the situation, thereby providing an optimal sales strategy for a physical store.

[0415] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0416] Step 1:

[0417] A user enters an assignment into a tablet

[0418] The user uses the tablet's user interface to input the problem they want to solve (e.g., "How to display a new product") in text format. The input problem text is saved on the device.

[0419] Input: User input text (issue)

[0420] Output: The assignment text is saved to your device

[0421] Step 2:

[0422] The device recognizes the user's emotions

[0423] While the user is entering the task, the device's camera captures the user's facial expressions and voice in real time, and an emotion recognition engine (EmotionRecognizer) is used to perform emotion analysis.

[0424] Input: User's facial expressions and voice data

[0425] Output: Emotional state data (e.g., tension, anxiety)

[0426] Step 3:

[0427] The device sends the task and emotion data to the server.

[0428] The task text and emotional state data are transmitted from the terminal to the server.

[0429] Input: Task text, emotional state data

[0430] Output: Data is sent to the server

[0431] Step 4:

[0432] The server parses the assignment text

[0433] The server uses natural language processing tools (NaiveBayesClassifier) ​​to analyze the submitted assignment text and extract key keywords and intent.

[0434] Input: Assignment text

[0435] Output: Analyzed data (keywords, task intent)

[0436] Step 5:

[0437] The server generates a virtual knowledge agent.

[0438] The server generates multiple virtual knowledge agents with different thinking patterns (e.g., sales maximization, profit maximization, cost minimization). Each agent analyzes the problem based on its own thinking pattern when it is generated.

[0439] Input: Analyzed data (keywords, problem intent)

[0440] Output: Multiple virtual knowledge agents

[0441] Step 6:

[0442] The server initiates a discussion with the virtual knowledge agent.

[0443] The virtual knowledge agents discuss issues from their own perspectives, taking into account their emotional states. During the discussion, emotional data is provided in real time, allowing the agents to adaptively adjust the content of the discussion based on this data.

[0444] Input: Multiple virtual knowledge agents, emotional state data

[0445] Output: Discussion results

[0446] Step 7:

[0447] The server aggregates the results of the discussions

[0448] The server aggregates the results of discussions among the virtual knowledge agents and derives an optimal conclusion, taking into account emotional data to arrive at a conclusion that is easily accepted by the user.

[0449] Input: Discussion results of virtual knowledge agents, emotional state data

[0450] Output: Aggregated optimal conclusion

[0451] Step 8:

[0452] The server sends the aggregated discussion results to the terminal.

[0453] The server sends the aggregated discussion results to the device, where users can review them and implement the proposed solutions.

[0454] Input: Aggregated discussion results

[0455] Output: The discussion results are sent to the terminal.

[0456] Step 9:

[0457] The device displays the best decision for the user.

[0458] The terminal displays the discussion results received from the server on the user interface. The user can review the proposed solutions (e.g., "The new product should be priced in the middle of the market and displayed at eye level") and use them to determine a sales strategy.

[0459] Input: Discussion results

[0460] Output: What is displayed to the user

[0461] 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.

[0462] 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.

[0463] 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.

[0464] [Second embodiment]

[0465] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0466] 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.

[0467] 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).

[0468] 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.

[0469] 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.

[0470] 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).

[0471] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0472] 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.

[0473] 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.

[0474] 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.

[0475] 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.

[0476] 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."

[0477] The present invention is a system in which a user inputs a problem through a terminal, and virtual knowledge agents with different thought patterns discuss the problem and derive an appropriate conclusion. The following describes in detail an embodiment of the present invention.

[0478] System Overview

[0479] 1. Enter your assignment

[0480] The user inputs the problem to be solved in text format using a terminal. For example, the user might input, "I want to determine the pricing strategy for new product A."

[0481] 2. Parsing the Input

[0482] The terminal receives the task input by the user and transmits it to the server.

[0483] The server analyzes the submitted text using natural language processing to extract key keywords and intent. Specifically, keywords such as "new product A," "pricing strategy," and "entering market" are extracted.

[0484] 3. Generation of virtual knowledge agents with diverse thinking patterns

[0485] The server generates virtual knowledge agents from an internal database, each with a different thought pattern.

[0486] Virtual Knowledge Agent 1: Thinking Patterns for Maximizing Sales

[0487] Virtual Knowledge Agent 2: Profit Maximization Thinking Pattern

[0488] Virtual Knowledge Agent 3: Cost Minimization Thinking Pattern

[0489] 4. Individual analysis of the issues

[0490] The server presents a task to each virtual knowledge agent and has them perform an analysis individually.

[0491] Virtual Knowledge Agent 1 considers the merits of a high-price strategy based on market data.

[0492] Virtual knowledge agent 2 considers the optimal profit margin based on revenue data and cost structure.

[0493] Virtual knowledge agent 3 analyzes strategies to minimize manufacturing and marketing costs.

[0494] 5. Initiating discussions and summarizing the results

[0495] The server initiates a discussion among the virtual knowledge agents on the topic, and each agent expresses their opinion from their own perspective and engages in the discussion.

[0496] The server aggregates these discussion results and derives the optimal conclusion. For example, the following discussion result is derived:

[0497] After considering the balance between whether to adopt a high-price strategy, a mid-price range, or a low-price strategy, it is determined that the mid-price range is most appropriate.

[0498] 6. Outputting the results

[0499] The server outputs the aggregated discussion results to the user's terminal through a user interface. As a specific example, the server may present to the user a conclusion such as setting the price of new product A in the mid-range to minimize manufacturing and marketing costs.

[0500] Specific examples

[0501] Consider the example of deciding on a pricing strategy for new product A. When a user types "I want to decide on a pricing strategy for new product A" into their terminal, the system proceeds according to the above procedure. Important keywords are extracted using natural language processing, and virtual knowledge agents discuss pricing strategies from their own perspectives. Ultimately, the server outputs the conclusion that "the price of new product A should be set in the mid-range, minimizing manufacturing and marketing costs," and this is displayed on the user's terminal.

[0502] This system allows appropriate conclusions to be reached through discussions from multiple perspectives, reducing the cost of discussions and improving work efficiency.

[0503] The processing flow will be explained below.

[0504] Step 1:

[0505] The user uses a terminal to input a problem in text format, such as "I want to decide on a pricing strategy for new product A."

[0506] Step 2:

[0507] The terminal receives the assignment text entered by the user and transmits this input data to the server.

[0508] Step 3:

[0509] To analyze the assignment text received by the server, natural language processing is used to perform text analysis. Specifically, key keywords and intents (e.g., "New Product A," "pricing strategy," "entering market," etc.) are extracted from the assignment text.

[0510] Step 4:

[0511] Based on the extracted keywords and intentions, the server generates virtual knowledge agents with different thought patterns from its internal database. Each agent is assigned the following thought patterns:

[0512] Virtual Knowledge Agent 1: Maximizing Sales

[0513] Virtual Knowledge Agent 2: Profit Maximization

[0514] Virtual Knowledge Agent 3: Cost Minimization

[0515] Step 5:

[0516] The server presents a task to each virtual knowledge agent and instructs them to perform an individual analysis. Each agent performs an individual analysis as follows:

[0517] Virtual Knowledge Agent 1 considers the merits of setting a higher price based on market data.

[0518] Virtual Knowledge Agent 2 considers the optimal profit margin based on revenue data and cost structure.

[0519] Virtual knowledge agent 3 considers strategies to minimize manufacturing and marketing costs.

[0520] Step 6:

[0521] The server starts a discussion among the virtual knowledge agents. This allows each agent to express their opinion from their own perspective. The specific flow of the exchange of opinions is as follows:

[0522] Virtual Knowledge Agent 1: "New Product A should be sold at a high price and targeted at the premium market to maximize sales."

[0523] Virtual Knowledge Agent 2: "To maximize profits, we should set the price at a reasonable level relative to our competitors' products."

[0524] Virtual Knowledge Agent 3: "We should adopt a low-price strategy when entering the market to keep our manufacturing and marketing costs low."

[0525] Step 7:

[0526] The server aggregates the results of the discussions among the virtual knowledge agents and derives an overall conclusion, such as "It is optimal to set the price of new product A in the mid-range and minimize manufacturing and marketing costs."

[0527] Step 8:

[0528] The server transmits the aggregated discussion results to the terminal through the user interface means, allowing the user to check the final conclusions and proposals.

[0529] Step 9:

[0530] The terminal displays the final conclusion sent from the server to the user. The user confirms the proposal, "Price new product A at a mid-range price to minimize manufacturing and marketing costs," and takes the proposal into consideration.

[0531] Through the above steps, the system of the present invention derives appropriate conclusions through discussions from a variety of perspectives, thereby improving business efficiency and reducing discussion costs.

[0532] Example 1

[0533] 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."

[0534] Conventional systems lack the ability to discuss the issues users face from multiple perspectives, making it difficult to reach optimal conclusions. In particular, the lack of virtual knowledge agents with different thinking patterns means that analysis is limited to a single perspective, which can lead to overlooking various approaches that may be useful in resolving the issue.

[0535] 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.

[0536] In this invention, the server includes a user interface means for inputting a problem, a natural language processing means for analyzing the input problem, a means for generating a plurality of virtual knowledge agents with different thought patterns, a means for presenting a problem to the virtual knowledge agents and having them analyze it individually, a means for initiating a discussion among the virtual knowledge agents and aggregating the results of the discussion, and a means for outputting the results of the discussion to the user interface means, thereby enabling users to reach an optimal conclusion through discussion from multiple perspectives.

[0537] A "problem" is a specific problem or theme that a user seeks to solve.

[0538] "User interface means" refers to the input / output interface used by the user to input tasks and display results.

[0539] "Natural language processing means" refers to technical means for analyzing natural language used by humans and extracting key keywords and intent.

[0540] A "virtual knowledge agent" is an artificial intelligence virtual agent that has a different thought pattern and can perform individual analysis of a problem.

[0541] A "discussion method" is a means by which multiple virtual knowledge agents exchange opinions from their own perspectives and carry out a process to arrive at an optimal conclusion.

[0542] The "aggregation means" is a technical means for collecting and integrating the discussion results of each virtual knowledge agent to derive a final conclusion.

[0543] "Output means" refers to the means used to present the results of the discussion to the user.

[0544] The present invention is a system in which a user inputs a problem through a terminal, and virtual knowledge agents with different thought patterns discuss the problem and derive an appropriate conclusion. An embodiment of this system will be described in detail below.

[0545] System Overview

[0546] The system includes the following main elements:

[0547] 1. User Interface Methods

[0548] The user inputs the task in text format. The terminal receives this input and provides an interface for sending it to the server. The user interface is implemented as a web browser or a dedicated application.

[0549] 2. Natural Language Processing Methods

[0550] The server receives the assignment text sent by the user and analyzes it using natural language processing techniques (e.g., Python's NLTK library). This analysis extracts key keywords and intent. For example, keywords such as "New Product A," "Pricing Strategy," and "Market Entry" are extracted.

[0551] 3. Creation of Virtual Knowledge Agents

[0552] The server generates virtual knowledge agents with different thought patterns from an internal database (e.g., MySQL), including different perspectives such as sales maximization, profit maximization, and cost minimization.

[0553] Virtual Knowledge Agent 1: Thinking Patterns for Maximizing Sales

[0554] Virtual Knowledge Agent 2: Profit Maximization Thinking Pattern

[0555] Virtual Knowledge Agent 3: Cost Minimization Thinking Pattern

[0556] 4. Individual analysis of the issues

[0557] The server presents each virtual knowledge agent with a task entered by the user and has them perform an individual analysis. Each agent performs the analysis based on relevant data sets (e.g., market research data, revenue data, manufacturing cost data).

[0558] 5. Initiating discussions and summarizing the results

[0559] The server initiates a discussion based on the analysis results of virtual knowledge agents with different thought patterns, and aggregates the results. This discussion is conducted via an API, sharing the data and perspectives of each agent to arrive at the optimal conclusion.

[0560] 6. Outputting the results

[0561] The server outputs the results of the discussion to the user interface and displays them on the user's device. For example, a conclusion such as "Price new product A should be set in the mid-range, minimizing manufacturing and marketing costs" may be presented.

[0562] Specific examples

[0563] Consider a scenario in which a pricing strategy for new product A is to be decided. A user inputs "I would like to decide on a pricing strategy for new product A" into their device. This problem is sent to a server, where important keywords are extracted using natural language processing technology. The server then generates virtual knowledge agents, each of which analyzes the problem from a different perspective. As a result, the agents debate each other, and the final conclusion is displayed on the user's device.

[0564] Prompt Sentence Examples

[0565] Below are some example prompts that can be input to a generative AI model:

[0566] "I'd like to decide on a pricing strategy for new product A. What are the advantages and disadvantages of a high-price strategy, a mid-price strategy, and a low-price strategy?"

[0567] "Please tell us the results of the discussions among the virtual knowledge agents regarding the marketing strategy for new product B."

[0568] The above prompts allow the user to understand how to operate the system and the output results through specific scenarios.

[0569] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0570] Step 1:

[0571] The user inputs the problem to be solved in text format into the terminal. By typing "I want to decide on the pricing strategy for new product A" in the input field and pressing the send button, the input problem is sent from the terminal to the server.

[0572] Step 2:

[0573] The terminal receives the assignment entered by the user and sends it to the server via an HTTP request. The server prepares the received assignment text data for analysis. The input data is in text format, and the server queues this data for analysis.

[0574] Step 3:

[0575] The server performs natural language processing on the received assignment text using Python's NLTK library. Here, the text is subjected to morphological analysis to extract key keywords and intent. This analysis extracts keywords such as "new product A," "pricing strategy," and "market entry." The output is a list of the extracted keywords.

[0576] Step 4:

[0577] The server generates virtual knowledge agents with different thinking patterns from an internal database (MySQL) based on the extracted keywords. Each agent is set with a thinking pattern of maximizing sales, maximizing profits, or minimizing costs. For example, virtual knowledge agent 1 has a thinking pattern of maximizing sales, while virtual knowledge agent 2 has a thinking pattern of maximizing profits. The output is instances of the multiple virtual knowledge agents generated.

[0578] Step 5:

[0579] The server presents the tasks entered by the user to the generated virtual knowledge agents and has them perform individual analyses. Each agent references related data sets (e.g., market research data or revenue data) and performs analysis based on its own thought patterns. For example, virtual knowledge agent 1 considers a high-price strategy based on market data. The output is a report of the analysis results by each agent.

[0580] Step 6:

[0581] The server initiates a discussion based on the analysis results of the virtual knowledge agents. In this discussion, the virtual knowledge agents exchange opinions via API and consider the issue from multiple perspectives. Each agent shares their analysis results and perspectives, and a discussion is held to arrive at the optimal conclusion. The output is the conclusion agreed upon through the discussion.

[0582] Step 7:

[0583] The server aggregates the results of the discussions and derives a final conclusion. For example, the conclusion may be, "The price of new product A should be set in the mid-range, minimizing manufacturing and marketing costs." This conclusion is prepared as text data.

[0584] Step 8:

[0585] The server sends the aggregated discussion results to the user interface, which displays them on the user's terminal. The user interface presents the results to the user in an easy-to-read format, for example, in the form of a dashboard or report. The output is the final conclusion displayed on the user's terminal.

[0586] By following the above steps, users can reach optimal conclusions through discussions from multiple perspectives.

[0587] (Application example 1)

[0588] 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."

[0589] Conventional content delivery systems have the problem of being unable to quickly and accurately recommend content that meets the detailed requirements of users. In particular, recommendations that take into account users' interests and viewing history require analysis and discussion from multiple perspectives. Therefore, an efficient and comprehensive method for providing optimal content to users is required.

[0590] 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.

[0591] In this invention, the server includes a user interface means for inputting a problem, a natural language processing means for analyzing the input problem, a means for generating a plurality of virtual knowledge agents with different thought patterns, a means for initiating a discussion among the virtual knowledge agents and aggregating the results of the discussion, a means for outputting the results of the discussion to the user interface means, a means for recommending content based on the problem input by the user, and a means for the virtual knowledge agents to analyze and evaluate the content from their own different perspectives. This makes it possible to recommend optimal content to users through comprehensive discussion and analysis from different perspectives.

[0592] "Issues" refer to problems or requests that users want to solve.

[0593] "User interface means" refers to an interface through which a user inputs tasks and receives results.

[0594] "Natural language processing means" refers to technical means that analyze input text and extract key keywords and intent.

[0595] A "virtual knowledge agent" refers to a virtual agent that has a specific thought pattern and analyzes and discusses issues.

[0596] "Discussion results" refers to the output obtained as the conclusion of a discussion conducted by multiple virtual knowledge agents.

[0597] "Emotion analysis" refers to the perspective of analyzing a user's emotions and psychological state.

[0598] "Trend following" refers to a perspective based on current trends and popularity.

[0599] "Data-driven" refers to a perspective based on a user's past data and viewing history.

[0600] "Means for recommending content" refers to means for recommending optimal content based on user input.

[0601] MODE FOR CARRYING OUT THE INVENTION

[0602] System Program

[0603] The system program for realizing this invention is a system in which virtual knowledge agents with different thought patterns recommend content based on the user's issues. This system is composed of a user interface means, natural language processing means, virtual knowledge agent generation means, discussion result aggregation means, and result output means.

[0604] Natural language explanation of the process

[0605] Hardware and Software Configuration

[0606] In this system, users input tasks using a smartphone or a head-mounted display (HMD). On the server side, we use the Hugging Face transformers library to perform natural language processing. The creation of virtual knowledge agents and the aggregation of discussions are implemented using Python code.

[0607] Data processing and calculation

[0608] 1. The user inputs a task via a smartphone or HMD. For example, the user inputs the text "I want to watch a relaxing movie."

[0609] 2. The device sends the input text to the server, which uses Hugging Face's transformers library for natural language processing to extract key keywords and intent.

[0610] 3. The server generates multiple virtual knowledge agents with different thinking patterns, including sentiment analysis, trend following, and data-driven perspectives.

[0611] 4. The virtual knowledge agents perform individual analyses of the input tasks from their own perspectives. For example, a sentiment analysis agent might recommend relaxing movies, while a trend-following agent might recommend the latest popular movies.

[0612] 5. The server aggregates the discussion results of each agent and generates an optimal content recommendation list.

[0613] 6. Finally, the results are output to the user via a smartphone or HMD, and the user can receive a list of recommended content.

[0614] Specific examples

[0615] As a concrete example, consider the case where a user enters "I want to watch a relaxing movie." In this case, the system will process it through the following steps:

[0616] User: "I want to watch a relaxing movie."

[0617] Emotion Agent: "This movie is relaxing."

[0618] Trend Agent: "Here are your latest favorite movies."

[0619] Data-Driven Agent: "I found some relaxing movies based on your viewing history."

[0620] In this way, optimal content is recommended to the user after comprehensive discussion from multiple perspectives.

[0621] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0622] Step 1:

[0623] The user inputs a task using a smartphone or a head-mounted display (HMD). For example, the user inputs text such as "I want to watch a relaxing movie." This input is sent to the system as a task.

[0624] Input: Assignment text "I want to watch a relaxing movie"

[0625] Output: Send the assignment text to the server

[0626] Step 2:

[0627] The terminal transmits the task text entered by the user to the server, which receives the text for natural language processing.

[0628] Input: Assignment text

[0629] Output: Assignment text received on the server

[0630] Step 3:

[0631] The server uses Hugging Face's transformers library to perform natural language analysis of the challenge text and extract key keywords and intent, such as "relax" and "movie."

[0632] Input: Assignment text

[0633] Output: Keywords "relax", "movie"

[0634] Step 4:

[0635] The server generates multiple virtual knowledge agents with different thinking patterns, such as sentiment analysis, trend following, and data-driven thinking.

[0636] Input:keyword

[0637] Output: Sentiment analysis agents, trend-following agents, data-driven agents

[0638] Step 5:

[0639] Each virtual knowledge agent performs a separate analysis of the input task from its own perspective. For example, a sentiment analysis agent recommends movies with a relaxing effect, a trend-following agent recommends the latest popular movies, and a data-driven agent recommends relaxing movies based on the user's viewing history.

[0640] Input:keyword

[0641] Output: Recommendation results of sentiment analysis agents, recommendation results of trend follower agents, recommendation results of data-driven agents

[0642] Step 6:

[0643] The server aggregates the discussion results of each agent and creates an optimal content recommendation list, which incorporates the recommendations of each agent.

[0644] Input: Recommendation results of each agent

[0645] Output: Best content recommendation list

[0646] Step 7:

[0647] Finally, the server outputs a list of optimal content recommendations to the user via their smartphone or HMD, allowing the user to select the most appropriate content from the list and watch it.

[0648] Input: Best content recommendation list

[0649] Output: Display a list of content recommendations to the user

[0650] 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.

[0651] The present invention is a system that derives more appropriate discussion results by combining an emotion engine that recognizes the emotional state of the user. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The following describes in detail the embodiments of the present invention.

[0652] System Overview

[0653] 1. Enter your assignment

[0654] The user uses a terminal to input a problem in text format, such as "I want to decide on a pricing strategy for new product A." At the same time, the user's emotional state is recognized.

[0655] 2. Parsing the Input

[0656] The terminal receives the task text and emotion data input by the user and transmits them to the server.

[0657] The server analyzes the submitted task text using natural language processing to extract key keywords and intent. Specifically, keywords such as "new product A," "pricing strategy," and "entering market" are extracted. Emotion data provided by the emotion engine is also taken into consideration.

[0658] 3. Generation of virtual knowledge agents with diverse thinking patterns

[0659] The server generates virtual knowledge agents from its internal database. Each agent is assigned a different thought pattern:

[0660] Virtual Knowledge Agent 1: Maximizing Sales

[0661] Virtual Knowledge Agent 2: Profit Maximization

[0662] Virtual Knowledge Agent 3: Cost Minimization

[0663] 4. Individual analysis of the issues

[0664] The server presents a task to each virtual knowledge agent and instructs them to perform an individual analysis. Each agent performs an individual analysis as follows:

[0665] Virtual Knowledge Agent 1 considers the merits of setting a higher price based on market data.

[0666] Virtual Knowledge Agent 2 considers the optimal profit margin based on revenue data and cost structure.

[0667] Virtual knowledge agent 3 considers strategies to minimize manufacturing and marketing costs.

[0668] 5. Initiating discussions and summarizing the results

[0669] The server initiates a discussion among the virtual knowledge agents, allowing each agent to express their own opinion from their own perspective. During the discussion, the emotion engine continuously monitors the user's emotional state and provides emotional data in real time. This allows the virtual knowledge agents to adaptively adjust the content of the discussion taking into account the emotional data.

[0670] The specific flow of the exchange of opinions is as follows:

[0671] Virtual Knowledge Agent 1: "New Product A should be sold at a high price and targeted at the premium market to maximize sales."

[0672] Virtual Knowledge Agent 2: "To maximize profits, we should set the price at a reasonable level relative to our competitors' products."

[0673] Virtual Knowledge Agent 3: "We should adopt a low-price strategy when entering the market to keep our manufacturing and marketing costs low."

[0674] 6. Aggregating discussion results and outputting them to the user

[0675] The server aggregates the results of the discussions and derives a comprehensive conclusion that takes into account sentiment data. For example, the following conclusion may be derived: "It is optimal to set the price of new product A in the mid-range and minimize manufacturing and marketing costs."

[0676] The server transmits the aggregated discussion results to the terminal through the user interface means, allowing the user to check the final conclusions and proposals.

[0677] 7. Displaying the results

[0678] The terminal displays the final conclusion sent from the server to the user. The user confirms the proposal, "Price new product A at a mid-range price to minimize manufacturing and marketing costs," and takes the proposal into consideration.

[0679] Specific examples

[0680] For example, if a user inputs a task to determine a pricing strategy for new product A, the system will proceed according to the procedure described above. In particular, if the user indicates emotions such as "tension" or "anxiety" when inputting, the emotion engine will detect this and the virtual knowledge agents will engage in discussions that will make the user feel more at ease, resulting in a conclusion that is easy for the user to accept. As a result, the optimal conclusion is provided to the user: "The price of new product A should be set in the middle range, while minimizing manufacturing and marketing costs."

[0681] This system uses an emotion engine to conduct discussions that take into account the user's emotional state, making it possible to derive more appropriate and practical conclusions.

[0682] The processing flow will be explained below.

[0683] Step 1:

[0684] The user uses the device to input a problem in text format, such as "I want to decide on a pricing strategy for new product A." At the same time, the user's emotional state is recognized from their facial expressions and voice.

[0685] Step 2:

[0686] The terminal receives the task text entered by the user and the emotion data analyzed by the emotion engine, and transmits them to the server.

[0687] Step 3:

[0688] To analyze the assignment text received by the server, natural language processing is used to perform text analysis. Specifically, key keywords and intent (e.g., "New Product A," "pricing strategy," "entering market," etc.) are extracted from the assignment text. At the same time, emotional data provided by the emotion engine is also taken into account in the analysis.

[0689] Step 4:

[0690] Based on the analysis results, the server generates virtual knowledge agents with different thought patterns from its internal database. Each agent is assigned the following thought patterns:

[0691] Virtual Knowledge Agent 1: Maximizing Sales

[0692] Virtual Knowledge Agent 2: Profit Maximization

[0693] Virtual Knowledge Agent 3: Cost Minimization

[0694] Step 5:

[0695] The server presents each virtual knowledge agent with a task and instructs them to perform an individual analysis. The specific analysis content is as follows:

[0696] Virtual Knowledge Agent 1 considers the merits of setting a higher price based on market data.

[0697] Virtual Knowledge Agent 2 considers the optimal profit margin based on revenue data and cost structure.

[0698] Virtual knowledge agent 3 considers strategies to minimize manufacturing and marketing costs.

[0699] Step 6:

[0700] The server initiates a discussion among the virtual knowledge agents. Each agent expresses their opinion from their own perspective as follows:

[0701] Virtual Knowledge Agent 1: "New Product A should be sold at a high price and targeted at the premium market to maximize sales."

[0702] Virtual Knowledge Agent 2: "To maximize profits, we should set the price at a reasonable level relative to our competitors' products."

[0703] Virtual Knowledge Agent 3: "We should adopt a low-price strategy when entering the market to keep our manufacturing and marketing costs low."

[0704] Step 7:

[0705] The server instructs the virtual knowledge agents to adaptively adjust the content of the discussion based on real-time emotional data provided by the emotion engine. For example, if the user is in a "tense" state, the agent will adjust the content of the discussion to make the user feel more at ease.

[0706] Step 8:

[0707] The server aggregates the results of the discussions and derives a comprehensive conclusion that takes into account sentiment data. For example, the following conclusion may be reached: "It is optimal to set the price of new product A in the mid-range and minimize manufacturing and marketing costs."

[0708] Step 9:

[0709] The server sends this conclusion to the terminal through the user interface means, allowing the user to check the final discussion results and conclusions.

[0710] Step 10:

[0711] The terminal displays the final conclusion sent from the server to the user, who confirms the proposal that "the price of new product A should be set in the mid-range to minimize manufacturing and marketing costs."

[0712] Through the detailed steps described above, the system of the present invention conducts discussions that take into account the user's emotional state, thereby improving work efficiency and deriving appropriate conclusions.

[0713] Example 2

[0714] 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."

[0715] Conventional systems were unable to consider the user's emotional state when deriving solutions to problems submitted by the user. As a result, even if practical and specific proposals were made, it was sometimes difficult to reach a conclusion that was easy for the user to accept. It was also difficult to derive a comprehensive conclusion through discussions from various perspectives. This has led to the need for a system that can provide the appropriate problem solutions that users desire.

[0716] 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.

[0717] In this invention, the server includes a user interface for inputting a task, a natural language processing unit for analyzing the input task and emotional data, a unit for generating a plurality of virtual knowledge agents with different thought patterns, a unit for initiating a discussion among the virtual knowledge agents, a unit for monitoring the emotional state of the user in real time during the discussion, and a unit for aggregating the results of the discussion, and a unit for outputting the results of the discussion to the user interface. This enables discussions that take the emotional state of the user into consideration, leading to more convincing conclusions. It is also possible to provide comprehensive proposals from the perspectives of a variety of virtual knowledge agents.

[0718] A "problem" is a specific problem or requirement that a user wants solved.

[0719] The "user interface means" is an interface through which a user inputs a task into the system.

[0720] "Natural language processing means" is a means for analyzing input text data and emotional data and extracting key keywords and intentions.

[0721] "Emotional data" is data that indicates the user's current emotional state and is based on information collected from facial expressions, voice, etc.

[0722] "Virtual Knowledge Agents" are virtual knowledge-based agents with different thinking patterns, each generated to analyze a problem according to a specific goal or policy.

[0723] "Discussion" is a process in which multiple virtual knowledge agents exchange opinions based on their own perspectives and seek optimal solutions.

[0724] "Discussion results" are the final conclusions or proposals derived from discussions among virtual knowledge agents.

[0725] The "means for outputting to the user interface means" is a means for allowing the user to visually confirm the results of the discussion.

[0726] "Real-time monitoring" refers to continuously observing the user's emotional state and instantly acquiring that data.

[0727] System Overview

[0728] The present invention is a system that derives more appropriate discussion results by combining an emotion engine that recognizes the emotional state of the user. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The following describes in detail the preferred embodiments of the present invention.

[0729] Hardware and Software Configuration

[0730] 1. Server

[0731] The server analyzes the input data using natural language processing tools (e.g., spaCy or NLTK) and generates a virtual knowledge agent using a generative AI model (e.g., GPT-4 or BERT) to facilitate the discussion.

[0732] The server is equipped with an emotion engine and is designed to monitor the user's emotional state in real time.

[0733] 2. Terminal

[0734] The terminal provides an interface for users to input tasks, and is equipped with features such as a text area, voice input, and a face recognition sensor.

[0735] The terminal is equipped with a communication module for transmitting the user's input data and emotion data to the server.

[0736] 3. Users

[0737] The user uses a terminal to input the issues and receive the review results.

[0738] The emotion engine recognizes the user's emotional state (e.g., "tension" or "anxiety") in real time.

[0739] Specific program behavior

[0740] 1. Task input and emotion recognition

[0741] The user uses the device to input a problem in text format, such as "I want to decide on a pricing strategy for new product A." At this time, sensors and cameras built into the device collect data on the user's emotions and analyze it in real time.

[0742] 2. Sending input data

[0743] The device sends the input text and emotion data to the server. The text data is packaged in JSON format, and the emotion data is encoded in a proprietary format.

[0744] 3. Analysis of input data

[0745] The server uses natural language processing tools to analyze the received text data and extract key keywords and intent. For example, keywords such as "new product A," "pricing strategy," and "entering market" are extracted. An emotion engine also interprets the emotional data and identifies states such as "tension" and "anxiety."

[0746] 4. Creation of Virtual Knowledge Agents

[0747] The server generates multiple virtual knowledge agents with different thinking patterns based on its internal database. For example, virtual knowledge agent 1 is set to maximize sales, virtual knowledge agent 2 to maximize profits, and virtual knowledge agent 3 to minimize costs.

[0748] Use generative AI models (e.g., GPT-4 or BERT) to build each agent's thought patterns.

[0749] 5. Performing individual analyses

[0750] The server requests each virtual knowledge agent to analyze a specific problem. For example, virtual knowledge agent 1 analyzes market data and considers the merits of setting high prices. Virtual knowledge agent 2 analyzes revenue data and cost structures to find the optimal profit margin. Virtual knowledge agent 3 analyzes manufacturing and marketing costs and proposes ways to minimize them.

[0751] 6. Initiating and Conducting a Discussion

[0752] The server initiates a discussion among the virtual knowledge agents, and each agent takes turns expressing their opinion. Virtual knowledge agent 1 states, "New product A should be sold at a high price and target the premium market," to which virtual knowledge agent 2 counters, "It should be priced at an appropriate level compared to competing products." During this time, the emotion engine continues to monitor the user's emotional state and provides data in real time.

[0753] 7. Summary of discussion results

[0754] The server aggregates the opinions of each agent and draws an overall conclusion taking into account the sentiment data. For example, it may conclude that "it is optimal to set the price of new product A in the mid-range and minimize manufacturing and marketing costs."

[0755] 8. Sending and displaying results to the user

[0756] The server sends the analysis results and conclusions of the discussion to the terminal via a user interface. The data is structured in JSON format and is displayed visually on the terminal.

[0757] The terminal displays the conclusion "The price of new product A should be set in the middle price range to minimize manufacturing and marketing costs" to the user, allowing the user to confirm the content.

[0758] Examples of concrete examples and prompts

[0759] For example, when a user inputs a task to determine a pricing strategy for new product A, the system proceeds according to the above procedure. In particular, if the user indicates emotions such as "tension" or "anxiety" when inputting, the emotion engine detects this, and the virtual knowledge agents engage in discussions that make the user feel more at ease, resulting in a conclusion that is easy for the user to accept. Ultimately, the optimal conclusion is provided to the user: "The price of new product A should be set in the middle range, minimizing manufacturing and marketing costs."

[0760] Example prompt sentence:

[0761] "Please suggest the optimal strategy to price New Product A in the mid-range and minimize manufacturing and marketing costs. Users are expressing uncertainty, so please provide a reassuring explanation."

[0762] This system uses an emotion engine to conduct discussions that take into account the user's emotional state, making it possible to derive more appropriate and practical conclusions.

[0763] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0764] Step 1:

[0765] User task input and emotion recognition

[0766] The user inputs a specific task (e.g., determining the pricing strategy for new product A) in text format through the terminal interface.

[0767] The device uses built-in sensors and cameras to collect emotional data in real time from the user's facial expressions, voice, etc.

[0768] Input: Task text, emotion data (facial expressions, voice, etc.)

[0769] Output: Prepare the combined task text and sentiment data as a data package for transmission.

[0770] How it works: The device receives user input, converts it into text format, and collects emotional data using sensors and compiles it into a single package.

[0771] Step 2:

[0772] Terminal sends input data

[0773] The device sends the combined task text and emotion data to the server, where the data is encoded in JSON or a dedicated format via a communication module.

[0774] Input: A data package containing the task text and sentiment data

[0775] Output: Data package sent to the server

[0776] How it works: The device sends data and the server receives it. Data transmission is designed to be highly reliable and with low latency.

[0777] Step 3:

[0778] Input analysis and emotional data acquisition by the server

[0779] The server parses the received data package. Natural language processing tools (e.g., spaCy or NLTK) are used to extract key keywords and intent from the assignment text. An emotion engine analyzes the emotion data and identifies the user's current emotional state (e.g., "nervous," "anxious," etc.).

[0780] Input: Issue text and sentiment data in JSON format

[0781] Output: Analysis results including extracted keywords and emotional states

[0782] How it works: Natural language processing tools perform text analysis, and the emotion engine processes emotion data in real time. For example, keywords such as "new product A" and "pricing strategy" are extracted, and the user's emotional state is identified as "anxiety."

[0783] Step 4:

[0784] Server-based virtual knowledge agent generation

[0785] The server generates multiple virtual knowledge agents with different thinking patterns (e.g., sales maximization, profit maximization, cost minimization) based on an internal database. Each agent is built using a generative AI model (e.g., GPT-4, BERT).

[0786] Input: Analysis results including extracted keywords and emotional states

[0787] Output: A list of virtual knowledge agents

[0788] Specific operation: Based on the analysis results, the server sets the thinking pattern of each agent and constructs the agent using an appropriate AI model.

[0789] Step 5:

[0790] Individual analysis performed by the server

[0791] The server requests each virtual knowledge agent to analyze a specific problem. For example, virtual knowledge agent 1 analyzes market data and considers the merits of setting a high price.

[0792] Input: List of virtual knowledge agents and analysis results

[0793] Output: The results of the analysis by each agent

[0794] Specific operation: Each agent retrieves the necessary information from the database and analyzes it based on its own thought pattern. For example, Agent 1 suggests, "New Product A should be sold at a high price and target the premium market."

[0795] Step 6:

[0796] Starting and running discussions by the server

[0797] The server initiates discussions among the virtual knowledge agents, and each agent expresses their opinion. The emotion engine monitors the user's emotional state in real time and provides the data.

[0798] Input: Analysis results and sentiment data for each agent

[0799] Output: Discussion progress and each agent's opinion

[0800] Specific operation: Agent 1 proposes a high price strategy, Agent 2 counters that the price should be set at an appropriate range, and Agent 3 emphasizes cost minimization. The emotion engine detects the user's "anxiety," and the agents adjust the content of the discussion based on that data.

[0801] Step 7:

[0802] Aggregation of discussion results by the server

[0803] The server aggregates the results of the discussions and derives an overall conclusion taking into account the emotional data.

[0804] Input: Discussion progress and sentiment data

[0805] Output: Overall conclusion

[0806] Specific operation: The server organizes the opinions of each agent and derives a conclusion, for example, that "it is optimal to set the price of new product A in the mid-range and minimize manufacturing and marketing costs."

[0807] Step 8:

[0808] The server sends the results to the user

[0809] The server sends the analysis results and conclusions of the discussion to the terminal via the user interface.

[0810] Input: Overall conclusion

[0811] Output: Conclusion data sent to the user terminal

[0812] Specific behavior: Data structured in JSON format is sent to the device.

[0813] Step 9:

[0814] Displaying results on a terminal

[0815] The terminal visually displays the results sent from the server to the user.

[0816] Input: Conclusion data sent from the server

[0817] Output: The final conclusion that is displayed to the user

[0818] Specific operation: The device displays the conclusion to the user, "New Product A should be priced in the mid-range to minimize manufacturing and marketing costs," and allows the user to confirm the conclusion.

[0819] (Application example 2)

[0820] 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."

[0821] When deciding on new product displays and pricing strategies in physical stores, employees and store managers face a variety of challenges, and there is a need to find more appropriate and effective solutions. However, decisions based on employees' individual experience and judgment are often influenced by emotions and intuition, making it difficult to derive optimal results. Furthermore, because it is not possible to take into account employees' emotional states, such as tension and anxiety, there is a lack of sufficient information to select the best strategy.

[0822] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0823] In this invention, the server includes a user interface means for inputting a problem, a natural language processing means for analyzing the input problem, a means for generating a plurality of virtual knowledge agents with different thought patterns, a means for initiating a discussion among the virtual knowledge agents and aggregating the discussion results, a means for outputting the discussion results to the user interface means, an emotion recognition means for recognizing the emotional state of the user, and a means for adaptively adjusting the discussion of the virtual knowledge agents based on the emotional state. This makes it possible to derive appropriate discussion results by the virtual knowledge agents while taking the emotional state of the user into consideration, and to provide optimal solutions for new product displays and pricing strategies in physical stores.

[0824] The "user interface means for inputting issues" refers to an interface that allows a user to input issues or requests that they wish to solve, and can be operated on the screen of a terminal.

[0825] The "natural language processing means for analyzing the input problem" is a means for analyzing the text input by the user and extracting the intention and main keywords.

[0826] "Multiple virtual knowledge agents with different thinking patterns" are virtual agents that analyze issues and offer opinions based on specific strategies or perspectives, and have different thinking patterns such as sales maximization, profit maximization, or cost minimization.

[0827] The "means for initiating a discussion among the virtual knowledge agents and summarizing the results of the discussion" is a means for a plurality of virtual knowledge agents to hold a discussion from their respective viewpoints and summarizing the results.

[0828] The "means for outputting the discussion results to the user interface means" refers to a means for displaying the summarized discussion results to the user.

[0829] The "emotion recognition means for recognizing the user's emotional state" is a means for recognizing the user's emotion from their facial expression or voice, and is used to understand the mental state of the user while they are inputting.

[0830] The "means for adaptively adjusting the discussion of the virtual knowledge agents based on the emotional state" refers to a means for adaptively adjusting the content and progress of the discussion of the virtual knowledge agents in consideration of the emotional state of the user.

[0831] To implement this invention, it is necessary to build a system that allows users to decide sales strategies for physical stores. This system includes the following components:

[0832] User Interface Means

[0833] The interface for users to input their tasks is a device such as a tablet or smartphone. This device provides a screen on which users can enter text about specific tasks, such as how to display a new product or a pricing strategy.

[0834] Natural language processing tools

[0835] To analyze the assignment text sent from the device, the server uses a natural language processing library (e.g., NaiveBayesClassifier), which allows it to extract key keywords and the intent of the assignment.

[0836] Virtual Knowledge Agent

[0837] The server generates multiple virtual knowledge agents with different thought patterns. For example, agents with perspectives such as maximizing sales, maximizing profits, and minimizing costs are generated. Each agent proposes the optimal strategy from its own perspective based on data analysis.

[0838] emotion recognition means

[0839] The device is equipped with a camera and uses an emotion recognition engine (e.g., EmotionRecognizer) to recognize the user's facial expressions and voice in real time, which makes it possible to understand the user's emotional state (tension, anxiety, etc.) while they are inputting.

[0840] Discussion and summary of results

[0841] The server instructs the virtual knowledge agents to hold discussions while taking into account the user's emotional state. The agents exchange opinions from different perspectives and ultimately derive an optimal conclusion. This result is displayed to the user through a user interface.

[0842] Specific examples

[0843] For example, if a user types "how to display new products" into the tablet, the system will function as follows:

[0844] 1. Task input: The user inputs "how to display a new product" on the tablet.

[0845] 2. Emotion Recognition: The emotion of "tension" is recognized through the camera while the user is typing.

[0846] 3. Natural Language Processing: The server analyzes the input text and extracts key keywords and intent.

[0847] 4. Virtual knowledge agent generation: Virtual knowledge agents with different thinking patterns are generated.

[0848] 5. Discussion and summary of results: The agents hold a discussion to reassure the nervous user and propose a strategy for displaying the product at eye level in the mid-price range.

[0849] 6. Displaying the results: The final proposal is displayed on the tablet, providing the user with specific advice such as "price the new product in the middle of the market and display it at eye level."

[0850] Prompt Sentence Examples

[0851] Example: Prompt text that the user types into the tablet

[0852] How to display new products

[0853] As described above, this invention recognizes the emotional state of the user and responds in real time, and virtual knowledge agents adaptively discuss the situation, thereby providing an optimal sales strategy for a physical store.

[0854] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0855] Step 1:

[0856] A user enters an assignment into a tablet

[0857] The user uses the tablet's user interface to input the problem they want to solve (e.g., "How to display a new product") in text format. The input problem text is saved on the device.

[0858] Input: User input text (issue)

[0859] Output: The assignment text is saved to your device

[0860] Step 2:

[0861] The device recognizes the user's emotions

[0862] While the user is entering the task, the device's camera captures the user's facial expressions and voice in real time, and an emotion recognition engine (EmotionRecognizer) is used to perform emotion analysis.

[0863] Input: User's facial expressions and voice data

[0864] Output: Emotional state data (e.g., tension, anxiety)

[0865] Step 3:

[0866] The device sends the task and emotion data to the server.

[0867] The task text and emotional state data are transmitted from the terminal to the server.

[0868] Input: Task text, emotional state data

[0869] Output: Data is sent to the server

[0870] Step 4:

[0871] The server parses the assignment text

[0872] The server uses natural language processing tools (NaiveBayesClassifier) ​​to analyze the submitted assignment text and extract key keywords and intent.

[0873] Input: Assignment text

[0874] Output: Analyzed data (keywords, task intent)

[0875] Step 5:

[0876] The server generates a virtual knowledge agent.

[0877] The server generates multiple virtual knowledge agents with different thinking patterns (e.g., sales maximization, profit maximization, cost minimization). Each agent analyzes the problem based on its own thinking pattern when it is generated.

[0878] Input: Analyzed data (keywords, problem intent)

[0879] Output: Multiple virtual knowledge agents

[0880] Step 6:

[0881] The server initiates a discussion with the virtual knowledge agent.

[0882] The virtual knowledge agents discuss issues from their own perspectives, taking into account their emotional states. During the discussion, emotional data is provided in real time, allowing the agents to adaptively adjust the content of the discussion based on this data.

[0883] Input: Multiple virtual knowledge agents, emotional state data

[0884] Output: Discussion results

[0885] Step 7:

[0886] The server aggregates the results of the discussions

[0887] The server aggregates the results of discussions among the virtual knowledge agents and derives an optimal conclusion, taking into account emotional data to arrive at a conclusion that is easily accepted by the user.

[0888] Input: Discussion results of virtual knowledge agents, emotional state data

[0889] Output: Aggregated optimal conclusion

[0890] Step 8:

[0891] The server sends the aggregated discussion results to the terminal.

[0892] The server sends the aggregated discussion results to the device, where users can review them and implement the proposed solutions.

[0893] Input: Aggregated discussion results

[0894] Output: The discussion results are sent to the terminal.

[0895] Step 9:

[0896] The device displays the best decision for the user.

[0897] The terminal displays the discussion results received from the server on the user interface. The user can review the proposed solutions (e.g., "The new product should be priced in the middle of the market and displayed at eye level") and use them to determine a sales strategy.

[0898] Input: Discussion results

[0899] Output: What is displayed to the user

[0900] 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.

[0901] 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.

[0902] 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.

[0903] [Third embodiment]

[0904] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0905] 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.

[0906] 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).

[0907] 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.

[0908] 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.

[0909] 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).

[0910] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0911] 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.

[0912] 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.

[0913] 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.

[0914] 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.

[0915] 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."

[0916] The present invention is a system in which a user inputs a problem through a terminal, and virtual knowledge agents with different thought patterns discuss the problem and derive an appropriate conclusion. The following describes in detail an embodiment of the present invention.

[0917] System Overview

[0918] 1. Enter your assignment

[0919] The user inputs the problem to be solved in text format using a terminal. For example, the user might input, "I want to determine the pricing strategy for new product A."

[0920] 2. Parsing the Input

[0921] The terminal receives the task input by the user and transmits it to the server.

[0922] The server analyzes the submitted text using natural language processing to extract key keywords and intent. Specifically, keywords such as "new product A," "pricing strategy," and "entering market" are extracted.

[0923] 3. Generation of virtual knowledge agents with diverse thinking patterns

[0924] The server generates virtual knowledge agents from an internal database, each with a different thought pattern.

[0925] Virtual Knowledge Agent 1: Thinking Patterns for Maximizing Sales

[0926] Virtual Knowledge Agent 2: Profit Maximization Thinking Pattern

[0927] Virtual Knowledge Agent 3: Cost Minimization Thinking Pattern

[0928] 4. Individual analysis of the issues

[0929] The server presents a task to each virtual knowledge agent and has them perform an analysis individually.

[0930] Virtual Knowledge Agent 1 considers the merits of a high-price strategy based on market data.

[0931] Virtual knowledge agent 2 considers the optimal profit margin based on revenue data and cost structure.

[0932] Virtual knowledge agent 3 analyzes strategies to minimize manufacturing and marketing costs.

[0933] 5. Initiating discussions and summarizing the results

[0934] The server initiates a discussion among the virtual knowledge agents on the topic, and each agent expresses their opinion from their own perspective and engages in the discussion.

[0935] The server aggregates these discussion results and derives the optimal conclusion. For example, the following discussion result is derived:

[0936] After considering the balance between whether to adopt a high-price strategy, a mid-price range, or a low-price strategy, it is determined that the mid-price range is most appropriate.

[0937] 6. Outputting the results

[0938] The server outputs the aggregated discussion results to the user's terminal through a user interface. As a specific example, the server may present to the user a conclusion such as setting the price of new product A in the mid-range to minimize manufacturing and marketing costs.

[0939] Specific examples

[0940] Consider the example of deciding on a pricing strategy for new product A. When a user types "I want to decide on a pricing strategy for new product A" into their terminal, the system proceeds according to the above procedure. Important keywords are extracted using natural language processing, and virtual knowledge agents discuss pricing strategies from their own perspectives. Ultimately, the server outputs the conclusion that "the price of new product A should be set in the mid-range, minimizing manufacturing and marketing costs," and this is displayed on the user's terminal.

[0941] This system allows appropriate conclusions to be reached through discussions from multiple perspectives, reducing the cost of discussions and improving work efficiency.

[0942] The processing flow will be explained below.

[0943] Step 1:

[0944] The user uses a terminal to input a problem in text format, such as "I want to decide on a pricing strategy for new product A."

[0945] Step 2:

[0946] The terminal receives the assignment text entered by the user and transmits this input data to the server.

[0947] Step 3:

[0948] To analyze the assignment text received by the server, natural language processing is used to perform text analysis. Specifically, key keywords and intents (e.g., "New Product A," "pricing strategy," "entering market," etc.) are extracted from the assignment text.

[0949] Step 4:

[0950] Based on the extracted keywords and intentions, the server generates virtual knowledge agents with different thought patterns from its internal database. Each agent is assigned the following thought patterns:

[0951] Virtual Knowledge Agent 1: Maximizing Sales

[0952] Virtual Knowledge Agent 2: Profit Maximization

[0953] Virtual Knowledge Agent 3: Cost Minimization

[0954] Step 5:

[0955] The server presents a task to each virtual knowledge agent and instructs them to perform an individual analysis. Each agent performs an individual analysis as follows:

[0956] Virtual Knowledge Agent 1 considers the merits of setting a higher price based on market data.

[0957] Virtual Knowledge Agent 2 considers the optimal profit margin based on revenue data and cost structure.

[0958] Virtual knowledge agent 3 considers strategies to minimize manufacturing and marketing costs.

[0959] Step 6:

[0960] The server starts a discussion among the virtual knowledge agents. This allows each agent to express their opinion from their own perspective. The specific flow of the exchange of opinions is as follows:

[0961] Virtual Knowledge Agent 1: "New Product A should be sold at a high price and targeted at the premium market to maximize sales."

[0962] Virtual Knowledge Agent 2: "To maximize profits, we should set the price at a reasonable level relative to our competitors' products."

[0963] Virtual Knowledge Agent 3: "We should adopt a low-price strategy when entering the market to keep our manufacturing and marketing costs low."

[0964] Step 7:

[0965] The server aggregates the results of the discussions among the virtual knowledge agents and derives an overall conclusion, such as "It is optimal to set the price of new product A in the mid-range and minimize manufacturing and marketing costs."

[0966] Step 8:

[0967] The server transmits the aggregated discussion results to the terminal through the user interface means, allowing the user to check the final conclusions and proposals.

[0968] Step 9:

[0969] The terminal displays the final conclusion sent from the server to the user. The user confirms the proposal, "Price new product A at a mid-range price to minimize manufacturing and marketing costs," and takes the proposal into consideration.

[0970] Through the above steps, the system of the present invention derives appropriate conclusions through discussions from a variety of perspectives, thereby improving business efficiency and reducing discussion costs.

[0971] Example 1

[0972] 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."

[0973] Conventional systems lack the ability to discuss the issues users face from multiple perspectives, making it difficult to reach optimal conclusions. In particular, the lack of virtual knowledge agents with different thinking patterns means that analysis is limited to a single perspective, which can lead to overlooking various approaches that may be useful in resolving the issue.

[0974] 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.

[0975] In this invention, the server includes a user interface means for inputting a problem, a natural language processing means for analyzing the input problem, a means for generating a plurality of virtual knowledge agents with different thought patterns, a means for presenting a problem to the virtual knowledge agents and having them analyze it individually, a means for initiating a discussion among the virtual knowledge agents and aggregating the results of the discussion, and a means for outputting the results of the discussion to the user interface means, thereby enabling users to reach an optimal conclusion through discussion from multiple perspectives.

[0976] A "problem" is a specific problem or theme that a user seeks to solve.

[0977] "User interface means" refers to the input / output interface used by the user to input tasks and display results.

[0978] "Natural language processing means" refers to technical means for analyzing natural language used by humans and extracting key keywords and intent.

[0979] A "virtual knowledge agent" is an artificial intelligence virtual agent that has a different thought pattern and can perform individual analysis of a problem.

[0980] A "discussion method" is a means by which multiple virtual knowledge agents exchange opinions from their own perspectives and carry out a process to arrive at an optimal conclusion.

[0981] The "aggregation means" is a technical means for collecting and integrating the discussion results of each virtual knowledge agent to derive a final conclusion.

[0982] "Output means" refers to the means used to present the results of the discussion to the user.

[0983] The present invention is a system in which a user inputs a problem through a terminal, and virtual knowledge agents with different thought patterns discuss the problem and derive an appropriate conclusion. An embodiment of this system will be described in detail below.

[0984] System Overview

[0985] The system includes the following main elements:

[0986] 1. User Interface Methods

[0987] The user inputs the task in text format. The terminal receives this input and provides an interface for sending it to the server. The user interface is implemented as a web browser or a dedicated application.

[0988] 2. Natural Language Processing Methods

[0989] The server receives the assignment text sent by the user and analyzes it using natural language processing techniques (e.g., Python's NLTK library). This analysis extracts key keywords and intent. For example, keywords such as "New Product A," "Pricing Strategy," and "Market Entry" are extracted.

[0990] 3. Creation of Virtual Knowledge Agents

[0991] The server generates virtual knowledge agents with different thought patterns from an internal database (e.g., MySQL), including different perspectives such as sales maximization, profit maximization, and cost minimization.

[0992] Virtual Knowledge Agent 1: Thinking Patterns for Maximizing Sales

[0993] Virtual Knowledge Agent 2: Profit Maximization Thinking Pattern

[0994] Virtual Knowledge Agent 3: Cost Minimization Thinking Pattern

[0995] 4. Individual analysis of the issues

[0996] The server presents each virtual knowledge agent with a task entered by the user and has them perform an individual analysis. Each agent performs the analysis based on relevant data sets (e.g., market research data, revenue data, manufacturing cost data).

[0997] 5. Initiating discussions and summarizing the results

[0998] The server initiates a discussion based on the analysis results of virtual knowledge agents with different thought patterns, and aggregates the results. This discussion is conducted via an API, sharing the data and perspectives of each agent to arrive at the optimal conclusion.

[0999] 6. Outputting the results

[1000] The server outputs the results of the discussion to the user interface and displays them on the user's device. For example, a conclusion such as "Price new product A should be set in the mid-range, minimizing manufacturing and marketing costs" may be presented.

[1001] Specific examples

[1002] Consider a scenario in which a pricing strategy for new product A is to be decided. A user inputs "I would like to decide on a pricing strategy for new product A" into their device. This problem is sent to a server, where important keywords are extracted using natural language processing technology. The server then generates virtual knowledge agents, each of which analyzes the problem from a different perspective. As a result, the agents debate each other, and the final conclusion is displayed on the user's device.

[1003] Prompt Sentence Examples

[1004] Below are some example prompts that can be input to a generative AI model:

[1005] "I'd like to decide on a pricing strategy for new product A. What are the advantages and disadvantages of a high-price strategy, a mid-price strategy, and a low-price strategy?"

[1006] "Please tell us the results of the discussions among the virtual knowledge agents regarding the marketing strategy for new product B."

[1007] The above prompts allow the user to understand how to operate the system and the output results through specific scenarios.

[1008] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1009] Step 1:

[1010] The user inputs the problem to be solved in text format into the terminal. By typing "I want to decide on the pricing strategy for new product A" in the input field and pressing the send button, the input problem is sent from the terminal to the server.

[1011] Step 2:

[1012] The terminal receives the assignment entered by the user and sends it to the server via an HTTP request. The server prepares the received assignment text data for analysis. The input data is in text format, and the server queues this data for analysis.

[1013] Step 3:

[1014] The server performs natural language processing on the received assignment text using Python's NLTK library. Here, the text is subjected to morphological analysis to extract key keywords and intent. This analysis extracts keywords such as "new product A," "pricing strategy," and "market entry." The output is a list of the extracted keywords.

[1015] Step 4:

[1016] The server generates virtual knowledge agents with different thinking patterns from an internal database (MySQL) based on the extracted keywords. Each agent is set with a thinking pattern of maximizing sales, maximizing profits, or minimizing costs. For example, virtual knowledge agent 1 has a thinking pattern of maximizing sales, while virtual knowledge agent 2 has a thinking pattern of maximizing profits. The output is instances of the multiple virtual knowledge agents generated.

[1017] Step 5:

[1018] The server presents the tasks entered by the user to the generated virtual knowledge agents and has them perform individual analyses. Each agent references related data sets (e.g., market research data or revenue data) and performs analysis based on its own thought patterns. For example, virtual knowledge agent 1 considers a high-price strategy based on market data. The output is a report of the analysis results by each agent.

[1019] Step 6:

[1020] The server initiates a discussion based on the analysis results of the virtual knowledge agents. In this discussion, the virtual knowledge agents exchange opinions via API and consider the issue from multiple perspectives. Each agent shares their analysis results and perspectives, and a discussion is held to arrive at the optimal conclusion. The output is the conclusion agreed upon through the discussion.

[1021] Step 7:

[1022] The server aggregates the results of the discussions and derives a final conclusion. For example, the conclusion may be, "The price of new product A should be set in the mid-range, minimizing manufacturing and marketing costs." This conclusion is prepared as text data.

[1023] Step 8:

[1024] The server sends the aggregated discussion results to the user interface, which displays them on the user's terminal. The user interface presents the results to the user in an easy-to-read format, for example, in the form of a dashboard or report. The output is the final conclusion displayed on the user's terminal.

[1025] By following the above steps, users can reach optimal conclusions through discussions from multiple perspectives.

[1026] (Application example 1)

[1027] 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."

[1028] Conventional content delivery systems have the problem of being unable to quickly and accurately recommend content that meets the detailed requirements of users. In particular, recommendations that take into account users' interests and viewing history require analysis and discussion from multiple perspectives. Therefore, an efficient and comprehensive method for providing optimal content to users is required.

[1029] 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.

[1030] In this invention, the server includes a user interface means for inputting a problem, a natural language processing means for analyzing the input problem, a means for generating a plurality of virtual knowledge agents with different thought patterns, a means for initiating a discussion among the virtual knowledge agents and aggregating the results of the discussion, a means for outputting the results of the discussion to the user interface means, a means for recommending content based on the problem input by the user, and a means for the virtual knowledge agents to analyze and evaluate the content from their own different perspectives. This makes it possible to recommend optimal content to users through comprehensive discussion and analysis from different perspectives.

[1031] "Issues" refer to problems or requests that users want to solve.

[1032] "User interface means" refers to an interface through which a user inputs tasks and receives results.

[1033] "Natural language processing means" refers to technical means that analyze input text and extract key keywords and intent.

[1034] A "virtual knowledge agent" refers to a virtual agent that has a specific thought pattern and analyzes and discusses issues.

[1035] "Discussion results" refers to the output obtained as the conclusion of a discussion conducted by multiple virtual knowledge agents.

[1036] "Emotion analysis" refers to the perspective of analyzing a user's emotions and psychological state.

[1037] "Trend following" refers to a perspective based on current trends and popularity.

[1038] "Data-driven" refers to a perspective based on a user's past data and viewing history.

[1039] "Means for recommending content" refers to means for recommending optimal content based on user input.

[1040] MODE FOR CARRYING OUT THE INVENTION

[1041] System Program

[1042] The system program for realizing this invention is a system in which virtual knowledge agents with different thought patterns recommend content based on the user's issues. This system is composed of a user interface means, natural language processing means, virtual knowledge agent generation means, discussion result aggregation means, and result output means.

[1043] Natural language explanation of the process

[1044] Hardware and Software Configuration

[1045] In this system, users input tasks using a smartphone or a head-mounted display (HMD). On the server side, we use the Hugging Face transformers library to perform natural language processing. The creation of virtual knowledge agents and the aggregation of discussions are implemented using Python code.

[1046] Data processing and calculation

[1047] 1. The user inputs a task via a smartphone or HMD. For example, the user inputs the text "I want to watch a relaxing movie."

[1048] 2. The device sends the input text to the server, which uses Hugging Face's transformers library for natural language processing to extract key keywords and intent.

[1049] 3. The server generates multiple virtual knowledge agents with different thinking patterns, including sentiment analysis, trend following, and data-driven perspectives.

[1050] 4. The virtual knowledge agents perform individual analyses of the input tasks from their own perspectives. For example, a sentiment analysis agent might recommend relaxing movies, while a trend-following agent might recommend the latest popular movies.

[1051] 5. The server aggregates the discussion results of each agent and generates an optimal content recommendation list.

[1052] 6. Finally, the results are output to the user via a smartphone or HMD, and the user can receive a list of recommended content.

[1053] Specific examples

[1054] As a concrete example, consider the case where a user enters "I want to watch a relaxing movie." In this case, the system will process it through the following steps:

[1055] User: "I want to watch a relaxing movie."

[1056] Emotion Agent: "This movie is relaxing."

[1057] Trend Agent: "Here are your latest favorite movies."

[1058] Data-Driven Agent: "I found some relaxing movies based on your viewing history."

[1059] In this way, optimal content is recommended to the user after comprehensive discussion from multiple perspectives.

[1060] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1061] Step 1:

[1062] The user inputs a task using a smartphone or a head-mounted display (HMD). For example, the user inputs text such as "I want to watch a relaxing movie." This input is sent to the system as a task.

[1063] Input: Assignment text "I want to watch a relaxing movie"

[1064] Output: Send the assignment text to the server

[1065] Step 2:

[1066] The terminal transmits the task text entered by the user to the server, which receives the text for natural language processing.

[1067] Input: Assignment text

[1068] Output: Assignment text received on the server

[1069] Step 3:

[1070] The server uses Hugging Face's transformers library to perform natural language analysis of the challenge text and extract key keywords and intent, such as "relax" and "movie."

[1071] Input: Assignment text

[1072] Output: Keywords "relax", "movie"

[1073] Step 4:

[1074] The server generates multiple virtual knowledge agents with different thinking patterns, such as sentiment analysis, trend following, and data-driven thinking.

[1075] Input:keyword

[1076] Output: Sentiment analysis agents, trend-following agents, data-driven agents

[1077] Step 5:

[1078] Each virtual knowledge agent performs a separate analysis of the input task from its own perspective. For example, a sentiment analysis agent recommends movies with a relaxing effect, a trend-following agent recommends the latest popular movies, and a data-driven agent recommends relaxing movies based on the user's viewing history.

[1079] Input:keyword

[1080] Output: Recommendation results of sentiment analysis agents, recommendation results of trend follower agents, recommendation results of data-driven agents

[1081] Step 6:

[1082] The server aggregates the discussion results of each agent and creates an optimal content recommendation list, which incorporates the recommendations of each agent.

[1083] Input: Recommendation results of each agent

[1084] Output: Best content recommendation list

[1085] Step 7:

[1086] Finally, the server outputs a list of optimal content recommendations to the user via their smartphone or HMD, allowing the user to select the most appropriate content from the list and watch it.

[1087] Input: Best content recommendation list

[1088] Output: Display a list of content recommendations to the user

[1089] 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.

[1090] The present invention is a system that derives more appropriate discussion results by combining an emotion engine that recognizes the emotional state of the user. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The following describes in detail the embodiments of the present invention.

[1091] System Overview

[1092] 1. Enter your assignment

[1093] The user uses a terminal to input a problem in text format, such as "I want to decide on a pricing strategy for new product A." At the same time, the user's emotional state is recognized.

[1094] 2. Parsing the Input

[1095] The terminal receives the task text and emotion data input by the user and transmits them to the server.

[1096] The server analyzes the submitted task text using natural language processing to extract key keywords and intent. Specifically, keywords such as "new product A," "pricing strategy," and "entering market" are extracted. Emotion data provided by the emotion engine is also taken into consideration.

[1097] 3. Generation of virtual knowledge agents with diverse thinking patterns

[1098] The server generates virtual knowledge agents from its internal database. Each agent is assigned a different thought pattern:

[1099] Virtual Knowledge Agent 1: Maximizing Sales

[1100] Virtual Knowledge Agent 2: Profit Maximization

[1101] Virtual Knowledge Agent 3: Cost Minimization

[1102] 4. Individual analysis of the issues

[1103] The server presents a task to each virtual knowledge agent and instructs them to perform an individual analysis. Each agent performs an individual analysis as follows:

[1104] Virtual Knowledge Agent 1 considers the merits of setting a higher price based on market data.

[1105] Virtual Knowledge Agent 2 considers the optimal profit margin based on revenue data and cost structure.

[1106] Virtual knowledge agent 3 considers strategies to minimize manufacturing and marketing costs.

[1107] 5. Initiating discussions and summarizing the results

[1108] The server initiates a discussion among the virtual knowledge agents, allowing each agent to express their own opinion from their own perspective. During the discussion, the emotion engine continuously monitors the user's emotional state and provides emotional data in real time. This allows the virtual knowledge agents to adaptively adjust the content of the discussion taking into account the emotional data.

[1109] The specific flow of the exchange of opinions is as follows:

[1110] Virtual Knowledge Agent 1: "New Product A should be sold at a high price and targeted at the premium market to maximize sales."

[1111] Virtual Knowledge Agent 2: "To maximize profits, we should set the price at a reasonable level relative to our competitors' products."

[1112] Virtual Knowledge Agent 3: "We should adopt a low-price strategy when entering the market to keep our manufacturing and marketing costs low."

[1113] 6. Aggregating discussion results and outputting them to the user

[1114] The server aggregates the results of the discussions and derives a comprehensive conclusion that takes into account sentiment data. For example, the following conclusion may be derived: "It is optimal to set the price of new product A in the mid-range and minimize manufacturing and marketing costs."

[1115] The server transmits the aggregated discussion results to the terminal through the user interface means, allowing the user to check the final conclusions and proposals.

[1116] 7. Displaying the results

[1117] The terminal displays the final conclusion sent from the server to the user. The user confirms the proposal, "Price new product A at a mid-range price to minimize manufacturing and marketing costs," and takes the proposal into consideration.

[1118] Specific examples

[1119] For example, if a user inputs a task to determine a pricing strategy for new product A, the system will proceed according to the procedure described above. In particular, if the user indicates emotions such as "tension" or "anxiety" when inputting, the emotion engine will detect this and the virtual knowledge agents will engage in discussions that will make the user feel more at ease, resulting in a conclusion that is easy for the user to accept. As a result, the optimal conclusion is provided to the user: "The price of new product A should be set in the middle range, while minimizing manufacturing and marketing costs."

[1120] This system uses an emotion engine to conduct discussions that take into account the user's emotional state, making it possible to derive more appropriate and practical conclusions.

[1121] The processing flow will be explained below.

[1122] Step 1:

[1123] The user uses the device to input a problem in text format, such as "I want to decide on a pricing strategy for new product A." At the same time, the user's emotional state is recognized from their facial expressions and voice.

[1124] Step 2:

[1125] The terminal receives the task text entered by the user and the emotion data analyzed by the emotion engine, and transmits them to the server.

[1126] Step 3:

[1127] To analyze the assignment text received by the server, natural language processing is used to perform text analysis. Specifically, key keywords and intent (e.g., "New Product A," "pricing strategy," "entering market," etc.) are extracted from the assignment text. At the same time, emotional data provided by the emotion engine is also taken into account in the analysis.

[1128] Step 4:

[1129] Based on the analysis results, the server generates virtual knowledge agents with different thought patterns from its internal database. Each agent is assigned the following thought patterns:

[1130] Virtual Knowledge Agent 1: Maximizing Sales

[1131] Virtual Knowledge Agent 2: Profit Maximization

[1132] Virtual Knowledge Agent 3: Cost Minimization

[1133] Step 5:

[1134] The server presents each virtual knowledge agent with a task and instructs them to perform an individual analysis. The specific analysis content is as follows:

[1135] Virtual Knowledge Agent 1 considers the merits of setting a higher price based on market data.

[1136] Virtual Knowledge Agent 2 considers the optimal profit margin based on revenue data and cost structure.

[1137] Virtual knowledge agent 3 considers strategies to minimize manufacturing and marketing costs.

[1138] Step 6:

[1139] The server initiates a discussion among the virtual knowledge agents. Each agent expresses their opinion from their own perspective as follows:

[1140] Virtual Knowledge Agent 1: "New Product A should be sold at a high price and targeted at the premium market to maximize sales."

[1141] Virtual Knowledge Agent 2: "To maximize profits, we should set the price at a reasonable level relative to our competitors' products."

[1142] Virtual Knowledge Agent 3: "We should adopt a low-price strategy when entering the market to keep our manufacturing and marketing costs low."

[1143] Step 7:

[1144] The server instructs the virtual knowledge agents to adaptively adjust the content of the discussion based on real-time emotional data provided by the emotion engine. For example, if the user is in a "tense" state, the agent will adjust the content of the discussion to make the user feel more at ease.

[1145] Step 8:

[1146] The server aggregates the results of the discussions and derives a comprehensive conclusion that takes into account sentiment data. For example, the following conclusion may be reached: "It is optimal to set the price of new product A in the mid-range and minimize manufacturing and marketing costs."

[1147] Step 9:

[1148] The server sends this conclusion to the terminal through the user interface means, allowing the user to check the final discussion results and conclusions.

[1149] Step 10:

[1150] The terminal displays the final conclusion sent from the server to the user, who confirms the proposal that "the price of new product A should be set in the mid-range to minimize manufacturing and marketing costs."

[1151] Through the detailed steps described above, the system of the present invention conducts discussions that take into account the user's emotional state, thereby improving work efficiency and deriving appropriate conclusions.

[1152] Example 2

[1153] 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."

[1154] Conventional systems were unable to consider the user's emotional state when deriving solutions to problems submitted by the user. As a result, even if practical and specific proposals were made, it was sometimes difficult to reach a conclusion that was easy for the user to accept. It was also difficult to derive a comprehensive conclusion through discussions from various perspectives. This has led to the need for a system that can provide the appropriate problem solutions that users desire.

[1155] 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.

[1156] In this invention, the server includes a user interface for inputting a task, a natural language processing unit for analyzing the input task and emotional data, a unit for generating a plurality of virtual knowledge agents with different thought patterns, a unit for initiating a discussion among the virtual knowledge agents, a unit for monitoring the user's emotional state in real time during the discussion, and a unit for aggregating the discussion results, and a unit for outputting the discussion results to the user interface. This enables discussions that take the user's emotional state into consideration, leading to more convincing conclusions. It is also possible to provide comprehensive proposals from the perspectives of a variety of virtual knowledge agents.

[1157] A "problem" is a specific problem or requirement that a user wants solved.

[1158] The "user interface means" is an interface through which a user inputs a task into the system.

[1159] "Natural language processing means" is a means for analyzing input text data and emotional data and extracting key keywords and intentions.

[1160] "Emotional data" is data that indicates the user's current emotional state and is based on information collected from facial expressions, voice, etc.

[1161] "Virtual Knowledge Agents" are virtual knowledge-based agents with different thinking patterns, each generated to analyze a problem according to a specific goal or policy.

[1162] "Discussion" is a process in which multiple virtual knowledge agents exchange opinions based on their own perspectives and seek optimal solutions.

[1163] "Discussion results" are the final conclusions or proposals derived from discussions among virtual knowledge agents.

[1164] The "means for outputting to the user interface means" is a means for allowing the user to visually confirm the results of the discussion.

[1165] "Real-time monitoring" refers to continuously observing the user's emotional state and instantly acquiring that data.

[1166] System Overview

[1167] The present invention is a system that derives more appropriate discussion results by combining an emotion engine that recognizes the emotional state of the user. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The following describes in detail the preferred embodiments of the present invention.

[1168] Hardware and Software Configuration

[1169] 1. Server

[1170] The server analyzes the input data using natural language processing tools (e.g., spaCy or NLTK) and generates a virtual knowledge agent using a generative AI model (e.g., GPT-4 or BERT) to facilitate the discussion.

[1171] The server is equipped with an emotion engine and is designed to monitor the user's emotional state in real time.

[1172] 2. Terminal

[1173] The terminal provides an interface for users to input tasks, and is equipped with features such as a text area, voice input, and a face recognition sensor.

[1174] The terminal is equipped with a communication module for transmitting the user's input data and emotion data to the server.

[1175] 3. Users

[1176] The user uses a terminal to input the issues and receive the review results.

[1177] The emotion engine recognizes the user's emotional state (e.g., "tension" or "anxiety") in real time.

[1178] Specific program behavior

[1179] 1. Task input and emotion recognition

[1180] The user uses the device to input a problem in text format, such as "I want to decide on a pricing strategy for new product A." At this time, sensors and cameras built into the device collect data on the user's emotions and analyze it in real time.

[1181] 2. Sending input data

[1182] The device sends the input text and emotion data to the server. The text data is packaged in JSON format, and the emotion data is encoded in a proprietary format.

[1183] 3. Analysis of input data

[1184] The server uses natural language processing tools to analyze the received text data and extract key keywords and intent. For example, keywords such as "new product A," "pricing strategy," and "entering market" are extracted. An emotion engine also interprets the emotional data and identifies states such as "tension" and "anxiety."

[1185] 4. Creation of Virtual Knowledge Agents

[1186] The server generates multiple virtual knowledge agents with different thinking patterns based on its internal database. For example, virtual knowledge agent 1 is set to maximize sales, virtual knowledge agent 2 to maximize profits, and virtual knowledge agent 3 to minimize costs.

[1187] Use generative AI models (e.g., GPT-4 or BERT) to build each agent's thought patterns.

[1188] 5. Performing individual analyses

[1189] The server requests each virtual knowledge agent to analyze a specific problem. For example, virtual knowledge agent 1 analyzes market data and considers the merits of setting high prices. Virtual knowledge agent 2 analyzes revenue data and cost structures to find the optimal profit margin. Virtual knowledge agent 3 analyzes manufacturing and marketing costs and proposes ways to minimize them.

[1190] 6. Initiating and Conducting a Discussion

[1191] The server initiates a discussion among the virtual knowledge agents, and each agent takes turns expressing their opinion. Virtual knowledge agent 1 states, "New product A should be sold at a high price and target the premium market," to which virtual knowledge agent 2 counters, "It should be priced at an appropriate level compared to competing products." During this time, the emotion engine continues to monitor the user's emotional state and provides data in real time.

[1192] 7. Summary of discussion results

[1193] The server aggregates the opinions of each agent and draws an overall conclusion taking into account the sentiment data. For example, it may conclude that "it is optimal to set the price of new product A in the mid-range and minimize manufacturing and marketing costs."

[1194] 8. Sending and displaying results to the user

[1195] The server sends the analysis results and conclusions of the discussion to the terminal via a user interface. The data is structured in JSON format and is displayed visually on the terminal.

[1196] The terminal displays the conclusion "The price of new product A should be set in the middle price range to minimize manufacturing and marketing costs" to the user, allowing the user to confirm the content.

[1197] Examples of concrete examples and prompts

[1198] For example, when a user inputs a task to determine a pricing strategy for new product A, the system proceeds according to the above procedure. In particular, if the user indicates emotions such as "tension" or "anxiety" when inputting, the emotion engine detects this, and the virtual knowledge agents engage in discussions that make the user feel more at ease, resulting in a conclusion that is easy for the user to accept. Ultimately, the optimal conclusion is provided to the user: "The price of new product A should be set in the middle range, minimizing manufacturing and marketing costs."

[1199] Example prompt sentence:

[1200] "Please suggest the optimal strategy to price New Product A in the mid-range and minimize manufacturing and marketing costs. Users are expressing uncertainty, so please provide a reassuring explanation."

[1201] This system uses an emotion engine to conduct discussions that take into account the user's emotional state, making it possible to derive more appropriate and practical conclusions.

[1202] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1203] Step 1:

[1204] User task input and emotion recognition

[1205] The user inputs a specific task (e.g., determining the pricing strategy for new product A) in text format through the terminal interface.

[1206] The device uses built-in sensors and cameras to collect emotional data in real time from the user's facial expressions, voice, etc.

[1207] Input: Task text, emotion data (facial expressions, voice, etc.)

[1208] Output: Prepare the combined task text and sentiment data as a data package for transmission.

[1209] How it works: The device receives user input, converts it into text format, and collects emotional data using sensors and compiles it into a single package.

[1210] Step 2:

[1211] Terminal sends input data

[1212] The device sends the combined task text and emotion data to the server, where the data is encoded in JSON or a dedicated format via a communication module.

[1213] Input: A data package containing the task text and sentiment data

[1214] Output: Data package sent to the server

[1215] How it works: The device sends data and the server receives it. Data transmission is designed to be highly reliable and with low latency.

[1216] Step 3:

[1217] Input analysis and emotional data acquisition by the server

[1218] The server parses the received data package. Natural language processing tools (e.g., spaCy or NLTK) are used to extract key keywords and intent from the assignment text. An emotion engine analyzes the emotion data and identifies the user's current emotional state (e.g., "nervous," "anxious," etc.).

[1219] Input: Issue text and sentiment data in JSON format

[1220] Output: Analysis results including extracted keywords and emotional states

[1221] How it works: Natural language processing tools perform text analysis, and the emotion engine processes emotion data in real time. For example, keywords such as "new product A" and "pricing strategy" are extracted, and the user's emotional state is identified as "anxiety."

[1222] Step 4:

[1223] Server-based virtual knowledge agent generation

[1224] The server generates multiple virtual knowledge agents with different thinking patterns (e.g., sales maximization, profit maximization, cost minimization) based on an internal database. Each agent is built using a generative AI model (e.g., GPT-4, BERT).

[1225] Input: Analysis results including extracted keywords and emotional states

[1226] Output: A list of virtual knowledge agents

[1227] Specific operation: Based on the analysis results, the server sets the thinking pattern of each agent and constructs the agent using an appropriate AI model.

[1228] Step 5:

[1229] Individual analysis performed by the server

[1230] The server requests each virtual knowledge agent to analyze a specific problem. For example, virtual knowledge agent 1 analyzes market data and considers the merits of setting a high price.

[1231] Input: List of virtual knowledge agents and analysis results

[1232] Output: The results of the analysis by each agent

[1233] Specific operation: Each agent retrieves the necessary information from the database and analyzes it based on its own thought pattern. For example, Agent 1 suggests, "New Product A should be sold at a high price and target the premium market."

[1234] Step 6:

[1235] Starting and running discussions by the server

[1236] The server initiates discussions among the virtual knowledge agents, and each agent expresses their opinion. The emotion engine monitors the user's emotional state in real time and provides the data.

[1237] Input: Analysis results and sentiment data for each agent

[1238] Output: Discussion progress and each agent's opinion

[1239] Specific operation: Agent 1 proposes a high price strategy, Agent 2 counters that the price should be set at an appropriate range, and Agent 3 emphasizes cost minimization. The emotion engine detects the user's "anxiety," and the agents adjust the content of the discussion based on that data.

[1240] Step 7:

[1241] Aggregation of discussion results by the server

[1242] The server aggregates the results of the discussions and derives an overall conclusion taking into account the emotional data.

[1243] Input: Discussion progress and sentiment data

[1244] Output: Overall conclusion

[1245] Specific operation: The server organizes the opinions of each agent and derives a conclusion, for example, that "it is optimal to set the price of new product A in the mid-range and minimize manufacturing and marketing costs."

[1246] Step 8:

[1247] The server sends the results to the user

[1248] The server sends the analysis results and conclusions of the discussion to the terminal via the user interface.

[1249] Input: Overall conclusion

[1250] Output: Conclusion data sent to the user terminal

[1251] Specific behavior: Data structured in JSON format is sent to the device.

[1252] Step 9:

[1253] Displaying results on a terminal

[1254] The terminal visually displays the results sent from the server to the user.

[1255] Input: Conclusion data sent from the server

[1256] Output: The final conclusion that is displayed to the user

[1257] Specific operation: The device displays the conclusion to the user, "New Product A should be priced in the mid-range to minimize manufacturing and marketing costs," and allows the user to confirm the conclusion.

[1258] (Application example 2)

[1259] 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."

[1260] When deciding on new product displays and pricing strategies in physical stores, employees and store managers face a variety of challenges, and there is a need to find more appropriate and effective solutions. However, decisions based on employees' individual experience and judgment are often influenced by emotions and intuition, making it difficult to derive optimal results. Furthermore, because it is not possible to take into account employees' emotional states, such as tension and anxiety, there is a lack of sufficient information to select the best strategy.

[1261] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1262] In this invention, the server includes a user interface means for inputting a problem, a natural language processing means for analyzing the input problem, a means for generating a plurality of virtual knowledge agents with different thought patterns, a means for initiating a discussion among the virtual knowledge agents and aggregating the discussion results, a means for outputting the discussion results to the user interface means, an emotion recognition means for recognizing the emotional state of the user, and a means for adaptively adjusting the discussion of the virtual knowledge agents based on the emotional state. This makes it possible to derive appropriate discussion results by the virtual knowledge agents while taking the emotional state of the user into consideration, and to provide optimal solutions for new product displays and pricing strategies in physical stores.

[1263] The "user interface means for inputting issues" refers to an interface that allows a user to input issues or requests that they wish to solve, and can be operated on the screen of a terminal.

[1264] The "natural language processing means for analyzing the input problem" is a means for analyzing the text input by the user and extracting the intention and main keywords.

[1265] "Multiple virtual knowledge agents with different thinking patterns" are virtual agents that analyze issues and offer opinions based on specific strategies or perspectives, and have different thinking patterns such as sales maximization, profit maximization, or cost minimization.

[1266] The "means for initiating a discussion among the virtual knowledge agents and summarizing the results of the discussion" is a means for a plurality of virtual knowledge agents to hold a discussion from their respective viewpoints and summarizing the results.

[1267] The "means for outputting the discussion results to the user interface means" refers to a means for displaying the summarized discussion results to the user.

[1268] The "emotion recognition means for recognizing the user's emotional state" is a means for recognizing the user's emotion from their facial expression or voice, and is used to understand the mental state of the user while they are inputting.

[1269] The "means for adaptively adjusting the discussion of the virtual knowledge agents based on the emotional state" refers to a means for adaptively adjusting the content and progress of the discussion of the virtual knowledge agents in consideration of the emotional state of the user.

[1270] To implement this invention, it is necessary to build a system that allows users to decide sales strategies for physical stores. This system includes the following components:

[1271] User Interface Means

[1272] The interface for users to input their tasks is a device such as a tablet or smartphone. This device provides a screen on which users can enter text about specific tasks, such as how to display a new product or a pricing strategy.

[1273] Natural language processing tools

[1274] To analyze the assignment text sent from the device, the server uses a natural language processing library (e.g., NaiveBayesClassifier), which allows it to extract key keywords and the intent of the assignment.

[1275] Virtual Knowledge Agent

[1276] The server generates multiple virtual knowledge agents with different thought patterns. For example, agents with perspectives such as maximizing sales, maximizing profits, and minimizing costs are generated. Each agent proposes the optimal strategy from its own perspective based on data analysis.

[1277] emotion recognition means

[1278] The device is equipped with a camera and uses an emotion recognition engine (e.g., EmotionRecognizer) to recognize the user's facial expressions and voice in real time, which makes it possible to understand the user's emotional state (tension, anxiety, etc.) while they are inputting.

[1279] Discussion and summary of results

[1280] The server instructs the virtual knowledge agents to hold discussions while taking into account the user's emotional state. The agents exchange opinions from different perspectives and ultimately derive an optimal conclusion. This result is displayed to the user through a user interface.

[1281] Specific examples

[1282] For example, if a user types "how to display new products" into the tablet, the system will function as follows:

[1283] 1. Task input: The user inputs "how to display a new product" on the tablet.

[1284] 2. Emotion Recognition: The emotion of "tension" is recognized through the camera while the user is typing.

[1285] 3. Natural Language Processing: The server analyzes the input text and extracts key keywords and intent.

[1286] 4. Virtual knowledge agent generation: Virtual knowledge agents with different thinking patterns are generated.

[1287] 5. Discussion and summary of results: The agents hold a discussion to reassure the nervous user and propose a strategy for displaying the product at eye level in the mid-price range.

[1288] 6. Displaying the results: The final proposal is displayed on the tablet, providing the user with specific advice such as "price the new product in the middle of the market and display it at eye level."

[1289] Prompt Sentence Examples

[1290] Example: Prompt text that the user types into the tablet

[1291] How to display new products

[1292] As described above, this invention recognizes the emotional state of the user and responds in real time, and virtual knowledge agents adaptively discuss the situation, thereby providing an optimal sales strategy for a physical store.

[1293] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1294] Step 1:

[1295] A user enters an assignment into a tablet

[1296] The user uses the tablet's user interface to input the problem they want to solve (e.g., "How to display a new product") in text format. The input problem text is saved on the device.

[1297] Input: User input text (issue)

[1298] Output: The assignment text is saved to your device

[1299] Step 2:

[1300] The device recognizes the user's emotions

[1301] While the user is entering the task, the device's camera captures the user's facial expressions and voice in real time, and an emotion recognition engine (EmotionRecognizer) is used to perform emotion analysis.

[1302] Input: User's facial expressions and voice data

[1303] Output: Emotional state data (e.g., tension, anxiety)

[1304] Step 3:

[1305] The device sends the task and emotion data to the server.

[1306] The task text and emotional state data are transmitted from the terminal to the server.

[1307] Input: Task text, emotional state data

[1308] Output: Data is sent to the server

[1309] Step 4:

[1310] The server parses the assignment text

[1311] The server uses natural language processing tools (NaiveBayesClassifier) ​​to analyze the submitted assignment text and extract key keywords and intent.

[1312] Input: Assignment text

[1313] Output: Analyzed data (keywords, task intent)

[1314] Step 5:

[1315] The server generates a virtual knowledge agent.

[1316] The server generates multiple virtual knowledge agents with different thinking patterns (e.g., sales maximization, profit maximization, cost minimization). Each agent analyzes the problem based on its own thinking pattern when it is generated.

[1317] Input: Analyzed data (keywords, problem intent)

[1318] Output: Multiple virtual knowledge agents

[1319] Step 6:

[1320] The server initiates a discussion with the virtual knowledge agent.

[1321] The virtual knowledge agents discuss issues from their own perspectives, taking into account their emotional states. During the discussion, emotional data is provided in real time, allowing the agents to adaptively adjust the content of the discussion based on this data.

[1322] Input: Multiple virtual knowledge agents, emotional state data

[1323] Output: Discussion results

[1324] Step 7:

[1325] The server aggregates the results of the discussions

[1326] The server aggregates the results of discussions among the virtual knowledge agents and derives an optimal conclusion, taking into account emotional data to arrive at a conclusion that is easily accepted by the user.

[1327] Input: Discussion results of virtual knowledge agents, emotional state data

[1328] Output: Aggregated optimal conclusion

[1329] Step 8:

[1330] The server sends the aggregated discussion results to the terminal.

[1331] The server sends the aggregated discussion results to the device, where users can review them and implement the proposed solutions.

[1332] Input: Aggregated discussion results

[1333] Output: The discussion results are sent to the terminal.

[1334] Step 9:

[1335] The device displays the best decision for the user.

[1336] The terminal displays the discussion results received from the server on the user interface. The user can review the proposed solutions (e.g., "The new product should be priced in the middle of the market and displayed at eye level") and use them to determine a sales strategy.

[1337] Input: Discussion results

[1338] Output: What is displayed to the user

[1339] 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.

[1340] 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.

[1341] 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.

[1342] [Fourth embodiment]

[1343] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1344] 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.

[1345] 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).

[1346] 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.

[1347] 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.

[1348] 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).

[1349] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1350] 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.

[1351] 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.

[1352] 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.

[1353] 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.

[1354] 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.

[1355] 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."

[1356] The present invention is a system in which a user inputs a problem through a terminal, and virtual knowledge agents with different thought patterns discuss the problem and derive an appropriate conclusion. The following describes in detail an embodiment of the present invention.

[1357] System Overview

[1358] 1. Enter your assignment

[1359] The user inputs the problem to be solved in text format using a terminal. For example, the user might input, "I want to determine the pricing strategy for new product A."

[1360] 2. Parsing the Input

[1361] The terminal receives the task input by the user and transmits it to the server.

[1362] The server analyzes the submitted text using natural language processing to extract key keywords and intent. Specifically, keywords such as "new product A," "pricing strategy," and "entering market" are extracted.

[1363] 3. Generation of virtual knowledge agents with diverse thinking patterns

[1364] The server generates virtual knowledge agents from an internal database, each with a different thought pattern.

[1365] Virtual Knowledge Agent 1: Thinking Patterns for Maximizing Sales

[1366] Virtual Knowledge Agent 2: Profit Maximization Thinking Pattern

[1367] Virtual Knowledge Agent 3: Cost Minimization Thinking Pattern

[1368] 4. Individual analysis of the issues

[1369] The server presents a task to each virtual knowledge agent and has them perform an analysis individually.

[1370] Virtual Knowledge Agent 1 considers the merits of a high-price strategy based on market data.

[1371] Virtual knowledge agent 2 considers the optimal profit margin based on revenue data and cost structure.

[1372] Virtual knowledge agent 3 analyzes strategies to minimize manufacturing and marketing costs.

[1373] 5. Initiating discussions and summarizing the results

[1374] The server initiates a discussion among the virtual knowledge agents on the topic, and each agent expresses their opinion from their own perspective and engages in the discussion.

[1375] The server aggregates these discussion results and derives the optimal conclusion. For example, the following discussion result is derived:

[1376] After considering the balance between whether to adopt a high-price strategy, a mid-price range, or a low-price strategy, it is determined that the mid-price range is most appropriate.

[1377] 6. Outputting the results

[1378] The server outputs the aggregated discussion results to the user's terminal through a user interface. As a specific example, the server may present to the user a conclusion such as setting the price of new product A in the mid-range to minimize manufacturing and marketing costs.

[1379] Specific examples

[1380] Consider the example of deciding on a pricing strategy for new product A. When a user types "I want to decide on a pricing strategy for new product A" into their terminal, the system proceeds according to the above procedure. Important keywords are extracted using natural language processing, and virtual knowledge agents discuss pricing strategies from their own perspectives. Ultimately, the server outputs the conclusion that "the price of new product A should be set in the mid-range, minimizing manufacturing and marketing costs," and this is displayed on the user's terminal.

[1381] This system allows appropriate conclusions to be reached through discussions from multiple perspectives, reducing the cost of discussions and improving work efficiency.

[1382] The processing flow will be explained below.

[1383] Step 1:

[1384] The user uses a terminal to input a problem in text format, such as "I want to decide on a pricing strategy for new product A."

[1385] Step 2:

[1386] The terminal receives the assignment text entered by the user and transmits this input data to the server.

[1387] Step 3:

[1388] To analyze the assignment text received by the server, natural language processing is used to perform text analysis. Specifically, key keywords and intents (e.g., "New Product A," "pricing strategy," "entering market," etc.) are extracted from the assignment text.

[1389] Step 4:

[1390] Based on the extracted keywords and intentions, the server generates virtual knowledge agents with different thought patterns from its internal database. Each agent is assigned the following thought patterns:

[1391] Virtual Knowledge Agent 1: Maximizing Sales

[1392] Virtual Knowledge Agent 2: Profit Maximization

[1393] Virtual Knowledge Agent 3: Cost Minimization

[1394] Step 5:

[1395] The server presents a task to each virtual knowledge agent and instructs them to perform an individual analysis. Each agent performs an individual analysis as follows:

[1396] Virtual Knowledge Agent 1 considers the merits of setting a higher price based on market data.

[1397] Virtual Knowledge Agent 2 considers the optimal profit margin based on revenue data and cost structure.

[1398] Virtual knowledge agent 3 considers strategies to minimize manufacturing and marketing costs.

[1399] Step 6:

[1400] The server starts a discussion among the virtual knowledge agents. This allows each agent to express their opinion from their own perspective. The specific flow of the exchange of opinions is as follows:

[1401] Virtual Knowledge Agent 1: "New Product A should be sold at a high price and targeted at the premium market to maximize sales."

[1402] Virtual Knowledge Agent 2: "To maximize profits, we should set the price at a reasonable level relative to our competitors' products."

[1403] Virtual Knowledge Agent 3: "We should adopt a low-price strategy when entering the market to keep our manufacturing and marketing costs low."

[1404] Step 7:

[1405] The server aggregates the results of the discussions among the virtual knowledge agents and derives an overall conclusion, such as "It is optimal to set the price of new product A in the mid-range and minimize manufacturing and marketing costs."

[1406] Step 8:

[1407] The server transmits the aggregated discussion results to the terminal through the user interface means, allowing the user to check the final conclusions and proposals.

[1408] Step 9:

[1409] The terminal displays the final conclusion sent from the server to the user. The user confirms the proposal, "Price new product A at a mid-range price to minimize manufacturing and marketing costs," and takes the proposal into consideration.

[1410] Through the above steps, the system of the present invention derives appropriate conclusions through discussions from a variety of perspectives, thereby improving business efficiency and reducing discussion costs.

[1411] Example 1

[1412] 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."

[1413] Conventional systems lack the ability to discuss the issues users face from multiple perspectives, making it difficult to reach optimal conclusions. In particular, the lack of virtual knowledge agents with different thinking patterns means that analysis is limited to a single perspective, which can lead to overlooking various approaches that may be useful in resolving the issue.

[1414] 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.

[1415] In this invention, the server includes a user interface means for inputting a problem, a natural language processing means for analyzing the input problem, a means for generating a plurality of virtual knowledge agents with different thought patterns, a means for presenting a problem to the virtual knowledge agents and having them analyze it individually, a means for initiating a discussion among the virtual knowledge agents and aggregating the results of the discussion, and a means for outputting the results of the discussion to the user interface means, thereby enabling users to reach an optimal conclusion through discussion from multiple perspectives.

[1416] A "problem" is a specific problem or theme that a user seeks to solve.

[1417] "User interface means" refers to the input / output interface used by the user to input tasks and display results.

[1418] "Natural language processing means" refers to technical means for analyzing natural language used by humans and extracting key keywords and intent.

[1419] A "virtual knowledge agent" is an artificial intelligence virtual agent that has a different thought pattern and can perform individual analysis of a problem.

[1420] A "discussion method" is a means by which multiple virtual knowledge agents exchange opinions from their own perspectives and carry out a process to arrive at an optimal conclusion.

[1421] The "aggregation means" is a technical means for collecting and integrating the discussion results of each virtual knowledge agent to derive a final conclusion.

[1422] "Output means" refers to the means used to present the results of the discussion to the user.

[1423] The present invention is a system in which a user inputs a problem through a terminal, and virtual knowledge agents with different thought patterns discuss the problem and derive an appropriate conclusion. An embodiment of this system will be described in detail below.

[1424] System Overview

[1425] The system includes the following main elements:

[1426] 1. User Interface Methods

[1427] The user inputs the task in text format. The terminal receives this input and provides an interface for sending it to the server. The user interface is implemented as a web browser or a dedicated application.

[1428] 2. Natural Language Processing Methods

[1429] The server receives the assignment text sent by the user and analyzes it using natural language processing techniques (e.g., Python's NLTK library). This analysis extracts key keywords and intent. For example, keywords such as "New Product A," "Pricing Strategy," and "Market Entry" are extracted.

[1430] 3. Creation of Virtual Knowledge Agents

[1431] The server generates virtual knowledge agents with different thought patterns from an internal database (e.g., MySQL), including different perspectives such as sales maximization, profit maximization, and cost minimization.

[1432] Virtual Knowledge Agent 1: Thinking Patterns for Maximizing Sales

[1433] Virtual Knowledge Agent 2: Profit Maximization Thinking Pattern

[1434] Virtual Knowledge Agent 3: Cost Minimization Thinking Pattern

[1435] 4. Individual analysis of the issues

[1436] The server presents each virtual knowledge agent with a task entered by the user and has them perform an individual analysis. Each agent performs the analysis based on relevant data sets (e.g., market research data, revenue data, manufacturing cost data).

[1437] 5. Initiating discussions and summarizing the results

[1438] The server initiates a discussion based on the analysis results of virtual knowledge agents with different thought patterns, and aggregates the results. This discussion is conducted via an API, sharing the data and perspectives of each agent to arrive at the optimal conclusion.

[1439] 6. Outputting the results

[1440] The server outputs the results of the discussion to the user interface and displays them on the user's device. For example, a conclusion such as "Price new product A should be set in the mid-range, minimizing manufacturing and marketing costs" may be presented.

[1441] Specific examples

[1442] Consider a scenario in which a pricing strategy for new product A is to be decided. A user inputs "I would like to decide on a pricing strategy for new product A" into their device. This problem is sent to a server, where important keywords are extracted using natural language processing technology. The server then generates virtual knowledge agents, each of which analyzes the problem from a different perspective. As a result, the agents debate each other, and the final conclusion is displayed on the user's device.

[1443] Prompt Sentence Examples

[1444] Below are some example prompts that can be input to a generative AI model:

[1445] "I'd like to decide on a pricing strategy for new product A. What are the advantages and disadvantages of a high-price strategy, a mid-price strategy, and a low-price strategy?"

[1446] "Please tell us the results of the discussions among the virtual knowledge agents regarding the marketing strategy for new product B."

[1447] The above prompts allow the user to understand how to operate the system and the output results through specific scenarios.

[1448] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1449] Step 1:

[1450] The user inputs the problem to be solved in text format into the terminal. By typing "I want to decide on the pricing strategy for new product A" in the input field and pressing the send button, the input problem is sent from the terminal to the server.

[1451] Step 2:

[1452] The terminal receives the assignment entered by the user and sends it to the server via an HTTP request. The server prepares the received assignment text data for analysis. The input data is in text format, and the server queues this data for analysis.

[1453] Step 3:

[1454] The server performs natural language processing on the received assignment text using Python's NLTK library. Here, the text is subjected to morphological analysis to extract key keywords and intent. This analysis extracts keywords such as "new product A," "pricing strategy," and "market entry." The output is a list of the extracted keywords.

[1455] Step 4:

[1456] The server generates virtual knowledge agents with different thinking patterns from an internal database (MySQL) based on the extracted keywords. Each agent is set with a thinking pattern of maximizing sales, maximizing profits, or minimizing costs. For example, virtual knowledge agent 1 has a thinking pattern of maximizing sales, while virtual knowledge agent 2 has a thinking pattern of maximizing profits. The output is instances of the multiple virtual knowledge agents generated.

[1457] Step 5:

[1458] The server presents the tasks entered by the user to the generated virtual knowledge agents and has them perform individual analyses. Each agent references related data sets (e.g., market research data or revenue data) and performs analysis based on its own thought patterns. For example, virtual knowledge agent 1 considers a high-price strategy based on market data. The output is a report of the analysis results by each agent.

[1459] Step 6:

[1460] The server initiates a discussion based on the analysis results of the virtual knowledge agents. In this discussion, the virtual knowledge agents exchange opinions via API and consider the issue from multiple perspectives. Each agent shares their analysis results and perspectives, and a discussion is held to arrive at the optimal conclusion. The output is the conclusion agreed upon through the discussion.

[1461] Step 7:

[1462] The server aggregates the results of the discussions and derives a final conclusion. For example, the conclusion may be, "The price of new product A should be set in the mid-range, minimizing manufacturing and marketing costs." This conclusion is prepared as text data.

[1463] Step 8:

[1464] The server sends the aggregated discussion results to the user interface, which displays them on the user's terminal. The user interface presents the results to the user in an easy-to-read format, for example, in the form of a dashboard or report. The output is the final conclusion displayed on the user's terminal.

[1465] By following the above steps, users can reach optimal conclusions through discussions from multiple perspectives.

[1466] (Application example 1)

[1467] 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."

[1468] Conventional content delivery systems have the problem of being unable to quickly and accurately recommend content that meets the detailed requirements of users. In particular, recommendations that take into account users' interests and viewing history require analysis and discussion from multiple perspectives. Therefore, an efficient and comprehensive method for providing optimal content to users is required.

[1469] 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.

[1470] In this invention, the server includes a user interface means for inputting a problem, a natural language processing means for analyzing the input problem, a means for generating a plurality of virtual knowledge agents with different thought patterns, a means for initiating a discussion among the virtual knowledge agents and aggregating the results of the discussion, a means for outputting the results of the discussion to the user interface means, a means for recommending content based on the problem input by the user, and a means for the virtual knowledge agents to analyze and evaluate the content from their own different perspectives. This makes it possible to recommend optimal content to users through comprehensive discussion and analysis from different perspectives.

[1471] "Issues" refer to problems or requests that users want to solve.

[1472] "User interface means" refers to an interface through which a user inputs tasks and receives results.

[1473] "Natural language processing means" refers to technical means that analyze input text and extract key keywords and intent.

[1474] A "virtual knowledge agent" refers to a virtual agent that has a specific thought pattern and analyzes and discusses issues.

[1475] "Discussion results" refers to the output obtained as the conclusion of a discussion conducted by multiple virtual knowledge agents.

[1476] "Emotion analysis" refers to the perspective of analyzing a user's emotions and psychological state.

[1477] "Trend following" refers to a perspective based on current trends and popularity.

[1478] "Data-driven" refers to a perspective based on a user's past data and viewing history.

[1479] "Means for recommending content" refers to means for recommending optimal content based on user input.

[1480] MODE FOR CARRYING OUT THE INVENTION

[1481] System Program

[1482] The system program for realizing this invention is a system in which virtual knowledge agents with different thought patterns recommend content based on the user's issues. This system is composed of a user interface means, natural language processing means, virtual knowledge agent generation means, discussion result aggregation means, and result output means.

[1483] Natural language explanation of the process

[1484] Hardware and Software Configuration

[1485] In this system, users input tasks using a smartphone or a head-mounted display (HMD). On the server side, we use the Hugging Face transformers library to perform natural language processing. The creation of virtual knowledge agents and the aggregation of discussions are implemented using Python code.

[1486] Data processing and calculation

[1487] 1. The user inputs a task via a smartphone or HMD. For example, the user inputs the text "I want to watch a relaxing movie."

[1488] 2. The device sends the input text to the server, which uses Hugging Face's transformers library for natural language processing to extract key keywords and intent.

[1489] 3. The server generates multiple virtual knowledge agents with different thinking patterns, including sentiment analysis, trend following, and data-driven perspectives.

[1490] 4. The virtual knowledge agents perform individual analyses of the input tasks from their own perspectives. For example, a sentiment analysis agent might recommend relaxing movies, while a trend-following agent might recommend the latest popular movies.

[1491] 5. The server aggregates the discussion results of each agent and generates an optimal content recommendation list.

[1492] 6. Finally, the results are output to the user via a smartphone or HMD, and the user can receive a list of recommended content.

[1493] Specific examples

[1494] As a concrete example, consider the case where a user enters "I want to watch a relaxing movie." In this case, the system will process it through the following steps:

[1495] User: "I want to watch a relaxing movie."

[1496] Emotion Agent: "This movie is relaxing."

[1497] Trend Agent: "Here are your latest favorite movies."

[1498] Data-Driven Agent: "I found some relaxing movies based on your viewing history."

[1499] In this way, optimal content is recommended to the user after comprehensive discussion from multiple perspectives.

[1500] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1501] Step 1:

[1502] The user inputs a task using a smartphone or a head-mounted display (HMD). For example, the user inputs text such as "I want to watch a relaxing movie." This input is sent to the system as a task.

[1503] Input: Assignment text "I want to watch a relaxing movie"

[1504] Output: Send the assignment text to the server

[1505] Step 2:

[1506] The terminal transmits the task text entered by the user to the server, which receives the text for natural language processing.

[1507] Input: Assignment text

[1508] Output: Assignment text received on the server

[1509] Step 3:

[1510] The server uses Hugging Face's transformers library to perform natural language analysis of the challenge text and extract key keywords and intent, such as "relax" and "movie."

[1511] Input: Assignment text

[1512] Output: Keywords "relax", "movie"

[1513] Step 4:

[1514] The server generates multiple virtual knowledge agents with different thinking patterns, such as sentiment analysis, trend following, and data-driven thinking.

[1515] Input:keyword

[1516] Output: Sentiment analysis agents, trend-following agents, data-driven agents

[1517] Step 5:

[1518] Each virtual knowledge agent performs a separate analysis of the input task from its own perspective. For example, a sentiment analysis agent recommends movies with a relaxing effect, a trend-following agent recommends the latest popular movies, and a data-driven agent recommends relaxing movies based on the user's viewing history.

[1519] Input:keyword

[1520] Output: Recommendation results of sentiment analysis agents, recommendation results of trend follower agents, recommendation results of data-driven agents

[1521] Step 6:

[1522] The server aggregates the discussion results of each agent and creates an optimal content recommendation list, which incorporates the recommendations of each agent.

[1523] Input: Recommendation results of each agent

[1524] Output: Best content recommendation list

[1525] Step 7:

[1526] Finally, the server outputs a list of optimal content recommendations to the user via their smartphone or HMD, allowing the user to select the most appropriate content from the list and watch it.

[1527] Input: Best content recommendation list

[1528] Output: Display a list of content recommendations to the user

[1529] 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.

[1530] The present invention is a system that derives more appropriate discussion results by combining an emotion engine that recognizes the emotional state of the user. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The following describes in detail the embodiments of the present invention.

[1531] System Overview

[1532] 1. Enter your assignment

[1533] The user uses a terminal to input a problem in text format, such as "I want to decide on a pricing strategy for new product A." At the same time, the user's emotional state is recognized.

[1534] 2. Parsing the Input

[1535] The terminal receives the task text and emotion data input by the user and transmits them to the server.

[1536] The server analyzes the submitted task text using natural language processing to extract key keywords and intent. Specifically, keywords such as "new product A," "pricing strategy," and "entering market" are extracted. Emotion data provided by the emotion engine is also taken into consideration.

[1537] 3. Generation of virtual knowledge agents with diverse thinking patterns

[1538] The server generates virtual knowledge agents from its internal database. Each agent is assigned a different thought pattern:

[1539] Virtual Knowledge Agent 1: Maximizing Sales

[1540] Virtual Knowledge Agent 2: Profit Maximization

[1541] Virtual Knowledge Agent 3: Cost Minimization

[1542] 4. Individual analysis of the issues

[1543] The server presents a task to each virtual knowledge agent and instructs them to perform an individual analysis. Each agent performs an individual analysis as follows:

[1544] Virtual Knowledge Agent 1 considers the merits of setting a higher price based on market data.

[1545] Virtual Knowledge Agent 2 considers the optimal profit margin based on revenue data and cost structure.

[1546] Virtual knowledge agent 3 considers strategies to minimize manufacturing and marketing costs.

[1547] 5. Initiating discussions and summarizing the results

[1548] The server initiates a discussion among the virtual knowledge agents, allowing each agent to express their own opinion from their own perspective. During the discussion, the emotion engine continuously monitors the user's emotional state and provides emotional data in real time. This allows the virtual knowledge agents to adaptively adjust the content of the discussion taking into account the emotional data.

[1549] The specific flow of the exchange of opinions is as follows:

[1550] Virtual Knowledge Agent 1: "New Product A should be sold at a high price and targeted at the premium market to maximize sales."

[1551] Virtual Knowledge Agent 2: "To maximize profits, we should set the price at a reasonable level relative to our competitors' products."

[1552] Virtual Knowledge Agent 3: "We should adopt a low-price strategy when entering the market to keep our manufacturing and marketing costs low."

[1553] 6. Aggregating discussion results and outputting them to the user

[1554] The server aggregates the results of the discussions and derives a comprehensive conclusion that takes into account sentiment data. For example, the following conclusion may be derived: "It is optimal to set the price of new product A in the mid-range and minimize manufacturing and marketing costs."

[1555] The server transmits the aggregated discussion results to the terminal through the user interface means, allowing the user to check the final conclusions and proposals.

[1556] 7. Displaying the results

[1557] The terminal displays the final conclusion sent from the server to the user. The user confirms the proposal, "Price new product A at a mid-range price to minimize manufacturing and marketing costs," and takes the proposal into consideration.

[1558] Specific examples

[1559] For example, if a user inputs a task to determine a pricing strategy for new product A, the system will proceed according to the procedure described above. In particular, if the user indicates emotions such as "tension" or "anxiety" when inputting, the emotion engine will detect this and the virtual knowledge agents will engage in discussions that will make the user feel more at ease, resulting in a conclusion that is easy for the user to accept. As a result, the optimal conclusion is provided to the user: "The price of new product A should be set in the middle range, while minimizing manufacturing and marketing costs."

[1560] This system uses an emotion engine to conduct discussions that take into account the user's emotional state, making it possible to derive more appropriate and practical conclusions.

[1561] The processing flow will be explained below.

[1562] Step 1:

[1563] The user uses the device to input a problem in text format, such as "I want to decide on a pricing strategy for new product A." At the same time, the user's emotional state is recognized from their facial expressions and voice.

[1564] Step 2:

[1565] The terminal receives the task text entered by the user and the emotion data analyzed by the emotion engine, and transmits them to the server.

[1566] Step 3:

[1567] To analyze the assignment text received by the server, natural language processing is used to perform text analysis. Specifically, key keywords and intent (e.g., "New Product A," "pricing strategy," "entering market," etc.) are extracted from the assignment text. At the same time, emotional data provided by the emotion engine is also taken into account in the analysis.

[1568] Step 4:

[1569] Based on the analysis results, the server generates virtual knowledge agents with different thought patterns from its internal database. Each agent is assigned the following thought patterns:

[1570] Virtual Knowledge Agent 1: Maximizing Sales

[1571] Virtual Knowledge Agent 2: Profit Maximization

[1572] Virtual Knowledge Agent 3: Cost Minimization

[1573] Step 5:

[1574] The server presents each virtual knowledge agent with a task and instructs them to perform an individual analysis. The specific analysis content is as follows:

[1575] Virtual Knowledge Agent 1 considers the merits of setting a higher price based on market data.

[1576] Virtual Knowledge Agent 2 considers the optimal profit margin based on revenue data and cost structure.

[1577] Virtual knowledge agent 3 considers strategies to minimize manufacturing and marketing costs.

[1578] Step 6:

[1579] The server initiates a discussion among the virtual knowledge agents. Each agent expresses their opinion from their own perspective as follows:

[1580] Virtual Knowledge Agent 1: "New Product A should be sold at a high price and targeted at the premium market to maximize sales."

[1581] Virtual Knowledge Agent 2: "To maximize profits, we should set the price at a reasonable level relative to our competitors' products."

[1582] Virtual Knowledge Agent 3: "We should adopt a low-price strategy when entering the market to keep our manufacturing and marketing costs low."

[1583] Step 7:

[1584] The server instructs the virtual knowledge agents to adaptively adjust the content of the discussion based on real-time emotional data provided by the emotion engine. For example, if the user is in a "tense" state, the agent will adjust the content of the discussion to make the user feel more at ease.

[1585] Step 8:

[1586] The server aggregates the results of the discussions and derives a comprehensive conclusion that takes into account sentiment data. For example, the following conclusion may be reached: "It is optimal to set the price of new product A in the mid-range and minimize manufacturing and marketing costs."

[1587] Step 9:

[1588] The server sends this conclusion to the terminal through the user interface means, allowing the user to check the final discussion results and conclusions.

[1589] Step 10:

[1590] The terminal displays the final conclusion sent from the server to the user, who confirms the proposal that "the price of new product A should be set in the mid-range to minimize manufacturing and marketing costs."

[1591] Through the detailed steps described above, the system of the present invention conducts discussions that take into account the user's emotional state, thereby improving work efficiency and deriving appropriate conclusions.

[1592] Example 2

[1593] 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."

[1594] Conventional systems were unable to consider the user's emotional state when deriving solutions to problems submitted by the user. As a result, even if practical and specific proposals were made, it was sometimes difficult to reach a conclusion that was easy for the user to accept. It was also difficult to derive a comprehensive conclusion through discussions from various perspectives. This has led to the need for a system that can provide the appropriate problem solutions that users desire.

[1595] 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.

[1596] In this invention, the server includes a user interface for inputting a task, a natural language processing unit for analyzing the input task and emotional data, a unit for generating a plurality of virtual knowledge agents with different thought patterns, a unit for initiating a discussion among the virtual knowledge agents, a unit for monitoring the emotional state of the user in real time during the discussion, and a unit for aggregating the results of the discussion, and a unit for outputting the results of the discussion to the user interface. This enables discussions that take the emotional state of the user into consideration, leading to more convincing conclusions. It is also possible to provide comprehensive proposals from the perspectives of a variety of virtual knowledge agents.

[1597] A "problem" is a specific problem or requirement that a user wants solved.

[1598] The "user interface means" is an interface through which a user inputs a task into the system.

[1599] "Natural language processing means" is a means for analyzing input text data and emotional data and extracting key keywords and intentions.

[1600] "Emotional data" is data that indicates the user's current emotional state and is based on information collected from facial expressions, voice, etc.

[1601] "Virtual Knowledge Agents" are virtual knowledge-based agents with different thinking patterns, each generated to analyze a problem according to a specific goal or policy.

[1602] "Discussion" is a process in which multiple virtual knowledge agents exchange opinions based on their own perspectives and seek optimal solutions.

[1603] "Discussion results" are the final conclusions or proposals derived from discussions among virtual knowledge agents.

[1604] The "means for outputting to the user interface means" is a means for allowing the user to visually confirm the results of the discussion.

[1605] "Real-time monitoring" refers to continuously observing the user's emotional state and instantly acquiring that data.

[1606] System Overview

[1607] The present invention is a system that derives more appropriate discussion results by combining an emotion engine that recognizes the emotional state of the user. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The following describes in detail the preferred embodiments of the present invention.

[1608] Hardware and Software Configuration

[1609] 1. Server

[1610] The server analyzes the input data using natural language processing tools (e.g., spaCy or NLTK) and generates a virtual knowledge agent using a generative AI model (e.g., GPT-4 or BERT) to facilitate the discussion.

[1611] The server is equipped with an emotion engine and is designed to monitor the user's emotional state in real time.

[1612] 2. Terminal

[1613] The terminal provides an interface for users to input tasks, and is equipped with features such as a text area, voice input, and a face recognition sensor.

[1614] The terminal is equipped with a communication module for transmitting the user's input data and emotion data to the server.

[1615] 3. Users

[1616] The user uses a terminal to input the issues and receive the review results.

[1617] The emotion engine recognizes the user's emotional state (e.g., "tension" or "anxiety") in real time.

[1618] Specific program behavior

[1619] 1. Task input and emotion recognition

[1620] The user uses the device to input a problem in text format, such as "I want to decide on a pricing strategy for new product A." At this time, sensors and cameras built into the device collect data on the user's emotions and analyze it in real time.

[1621] 2. Sending input data

[1622] The device sends the input text and emotion data to the server. The text data is packaged in JSON format, and the emotion data is encoded in a proprietary format.

[1623] 3. Analysis of input data

[1624] The server uses natural language processing tools to analyze the received text data and extract key keywords and intent. For example, keywords such as "new product A," "pricing strategy," and "entering market" are extracted. An emotion engine also interprets the emotional data and identifies states such as "tension" and "anxiety."

[1625] 4. Creation of Virtual Knowledge Agents

[1626] The server generates multiple virtual knowledge agents with different thinking patterns based on its internal database. For example, virtual knowledge agent 1 is set to maximize sales, virtual knowledge agent 2 to maximize profits, and virtual knowledge agent 3 to minimize costs.

[1627] Use generative AI models (e.g., GPT-4 or BERT) to build each agent's thought patterns.

[1628] 5. Performing individual analyses

[1629] The server requests each virtual knowledge agent to analyze a specific problem. For example, virtual knowledge agent 1 analyzes market data and considers the merits of setting high prices. Virtual knowledge agent 2 analyzes revenue data and cost structures to find the optimal profit margin. Virtual knowledge agent 3 analyzes manufacturing and marketing costs and proposes ways to minimize them.

[1630] 6. Initiating and Conducting a Discussion

[1631] The server initiates a discussion among the virtual knowledge agents, and each agent takes turns expressing their opinion. Virtual knowledge agent 1 states, "New product A should be sold at a high price and target the premium market," to which virtual knowledge agent 2 counters, "It should be priced at an appropriate level compared to competing products." During this time, the emotion engine continues to monitor the user's emotional state and provides data in real time.

[1632] 7. Summary of discussion results

[1633] The server aggregates the opinions of each agent and draws an overall conclusion taking into account the sentiment data. For example, it may conclude that "it is optimal to set the price of new product A in the mid-range and minimize manufacturing and marketing costs."

[1634] 8. Sending and displaying results to the user

[1635] The server sends the analysis results and conclusions of the discussion to the terminal via a user interface. The data is structured in JSON format and is displayed visually on the terminal.

[1636] The terminal displays the conclusion "The price of new product A should be set in the middle price range to minimize manufacturing and marketing costs" to the user, allowing the user to confirm the content.

[1637] Examples of concrete examples and prompts

[1638] For example, when a user inputs a task to determine a pricing strategy for new product A, the system proceeds according to the above procedure. In particular, if the user indicates emotions such as "tension" or "anxiety" when inputting, the emotion engine detects this, and the virtual knowledge agents engage in discussions that make the user feel more at ease, resulting in a conclusion that is easy for the user to accept. Ultimately, the optimal conclusion is provided to the user: "The price of new product A should be set in the middle range, minimizing manufacturing and marketing costs."

[1639] Example prompt sentence:

[1640] "Please suggest the optimal strategy to price New Product A in the mid-range and minimize manufacturing and marketing costs. Users are expressing uncertainty, so please provide a reassuring explanation."

[1641] This system uses an emotion engine to conduct discussions that take into account the user's emotional state, making it possible to derive more appropriate and practical conclusions.

[1642] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1643] Step 1:

[1644] User task input and emotion recognition

[1645] The user inputs a specific task (e.g., determining the pricing strategy for new product A) in text format through the terminal interface.

[1646] The device uses built-in sensors and cameras to collect emotional data in real time from the user's facial expressions, voice, etc.

[1647] Input: Task text, emotion data (facial expressions, voice, etc.)

[1648] Output: Prepare the combined task text and sentiment data as a data package for transmission.

[1649] How it works: The device receives user input, converts it into text format, and collects emotional data using sensors and compiles it into a single package.

[1650] Step 2:

[1651] Terminal sends input data

[1652] The device sends the combined task text and emotion data to the server, where the data is encoded in JSON or a dedicated format via a communication module.

[1653] Input: A data package containing the task text and sentiment data

[1654] Output: Data package sent to the server

[1655] How it works: The device sends data and the server receives it. Data transmission is designed to be highly reliable and with low latency.

[1656] Step 3:

[1657] Input analysis and emotional data acquisition by the server

[1658] The server parses the received data package. Natural language processing tools (e.g., spaCy or NLTK) are used to extract key keywords and intent from the assignment text. An emotion engine analyzes the emotion data and identifies the user's current emotional state (e.g., "nervous," "anxious," etc.).

[1659] Input: Issue text and sentiment data in JSON format

[1660] Output: Analysis results including extracted keywords and emotional states

[1661] How it works: Natural language processing tools perform text analysis, and the emotion engine processes emotion data in real time. For example, keywords such as "new product A" and "pricing strategy" are extracted, and the user's emotional state is identified as "anxiety."

[1662] Step 4:

[1663] Server-based virtual knowledge agent generation

[1664] The server generates multiple virtual knowledge agents with different thinking patterns (e.g., sales maximization, profit maximization, cost minimization) based on an internal database. Each agent is built using a generative AI model (e.g., GPT-4, BERT).

[1665] Input: Analysis results including extracted keywords and emotional states

[1666] Output: A list of virtual knowledge agents

[1667] Specific operation: Based on the analysis results, the server sets the thinking pattern of each agent and constructs the agent using an appropriate AI model.

[1668] Step 5:

[1669] Individual analysis performed by the server

[1670] The server requests each virtual knowledge agent to analyze a specific problem. For example, virtual knowledge agent 1 analyzes market data and considers the merits of setting a high price.

[1671] Input: List of virtual knowledge agents and analysis results

[1672] Output: The results of the analysis by each agent

[1673] Specific operation: Each agent retrieves the necessary information from the database and analyzes it based on its own thought pattern. For example, Agent 1 suggests, "New Product A should be sold at a high price and target the premium market."

[1674] Step 6:

[1675] Starting and running discussions by the server

[1676] The server initiates discussions among the virtual knowledge agents, and each agent expresses their opinion. The emotion engine monitors the user's emotional state in real time and provides the data.

[1677] Input: Analysis results and sentiment data for each agent

[1678] Output: Discussion progress and each agent's opinion

[1679] Specific operation: Agent 1 proposes a high price strategy, Agent 2 counters that the price should be set at an appropriate range, and Agent 3 emphasizes cost minimization. The emotion engine detects the user's "anxiety," and the agents adjust the content of the discussion based on that data.

[1680] Step 7:

[1681] Aggregation of discussion results by the server

[1682] The server aggregates the results of the discussions and derives an overall conclusion taking into account the emotional data.

[1683] Input: Discussion progress and sentiment data

[1684] Output: Overall conclusion

[1685] Specific operation: The server organizes the opinions of each agent and derives a conclusion, for example, that "it is optimal to set the price of new product A in the mid-range and minimize manufacturing and marketing costs."

[1686] Step 8:

[1687] The server sends the results to the user

[1688] The server sends the analysis results and conclusions of the discussion to the terminal via the user interface.

[1689] Input: Overall conclusion

[1690] Output: Conclusion data sent to the user terminal

[1691] Specific behavior: Data structured in JSON format is sent to the device.

[1692] Step 9:

[1693] Displaying results on a terminal

[1694] The terminal visually displays the results sent from the server to the user.

[1695] Input: Conclusion data sent from the server

[1696] Output: The final conclusion that is displayed to the user

[1697] Specific operation: The device displays the conclusion to the user, "New Product A should be priced in the mid-range to minimize manufacturing and marketing costs," and allows the user to confirm the conclusion.

[1698] (Application example 2)

[1699] 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."

[1700] When deciding on new product displays and pricing strategies in physical stores, employees and store managers face a variety of challenges, and there is a need to find more appropriate and effective solutions. However, decisions based on employees' individual experience and judgment are often influenced by emotions and intuition, making it difficult to derive optimal results. Furthermore, because it is not possible to take into account employees' emotional states, such as tension and anxiety, there is a lack of sufficient information to select the best strategy.

[1701] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1702] In this invention, the server includes a user interface means for inputting a problem, a natural language processing means for analyzing the input problem, a means for generating a plurality of virtual knowledge agents with different thought patterns, a means for initiating a discussion among the virtual knowledge agents and aggregating the discussion results, a means for outputting the discussion results to the user interface means, an emotion recognition means for recognizing the emotional state of the user, and a means for adaptively adjusting the discussion of the virtual knowledge agents based on the emotional state. This makes it possible to derive appropriate discussion results by the virtual knowledge agents while taking the emotional state of the user into consideration, and to provide optimal solutions for new product displays and pricing strategies in physical stores.

[1703] The "user interface means for inputting issues" refers to an interface that allows a user to input issues or requests that they wish to solve, and can be operated on the screen of a terminal.

[1704] The "natural language processing means for analyzing the input problem" is a means for analyzing the text input by the user and extracting the intention and main keywords.

[1705] "Multiple virtual knowledge agents with different thinking patterns" are virtual agents that analyze issues and offer opinions based on specific strategies or perspectives, and have different thinking patterns such as sales maximization, profit maximization, or cost minimization.

[1706] The "means for initiating a discussion among the virtual knowledge agents and summarizing the results of the discussion" is a means for a plurality of virtual knowledge agents to hold a discussion from their respective viewpoints and summarizing the results.

[1707] The "means for outputting the discussion results to the user interface means" refers to a means for displaying the summarized discussion results to the user.

[1708] The "emotion recognition means for recognizing the user's emotional state" is a means for recognizing the user's emotion from their facial expression or voice, and is used to understand the mental state of the user while they are inputting.

[1709] The "means for adaptively adjusting the discussion of the virtual knowledge agents based on the emotional state" refers to a means for adaptively adjusting the content and progress of the discussion of the virtual knowledge agents in consideration of the emotional state of the user.

[1710] To implement this invention, it is necessary to build a system that allows users to decide sales strategies for physical stores. This system includes the following components:

[1711] User Interface Means

[1712] The interface for users to input their tasks is a device such as a tablet or smartphone. This device provides a screen on which users can enter text about specific tasks, such as how to display a new product or a pricing strategy.

[1713] Natural language processing tools

[1714] To analyze the assignment text sent from the device, the server uses a natural language processing library (e.g., NaiveBayesClassifier), which allows it to extract key keywords and the intent of the assignment.

[1715] Virtual Knowledge Agent

[1716] The server generates multiple virtual knowledge agents with different thought patterns. For example, agents with perspectives such as maximizing sales, maximizing profits, and minimizing costs are generated. Each agent proposes the optimal strategy from its own perspective based on data analysis.

[1717] emotion recognition means

[1718] The device is equipped with a camera and uses an emotion recognition engine (e.g., EmotionRecognizer) to recognize the user's facial expressions and voice in real time, which makes it possible to understand the user's emotional state (tension, anxiety, etc.) while they are inputting.

[1719] Discussion and summary of results

[1720] The server instructs the virtual knowledge agents to hold discussions while taking into account the user's emotional state. The agents exchange opinions from different perspectives and ultimately derive an optimal conclusion. This result is displayed to the user through a user interface.

[1721] Specific examples

[1722] For example, if a user types "how to display new products" into the tablet, the system will function as follows:

[1723] 1. Task input: The user inputs "how to display a new product" on the tablet.

[1724] 2. Emotion Recognition: The emotion of "tension" is recognized through the camera while the user is typing.

[1725] 3. Natural Language Processing: The server analyzes the input text and extracts key keywords and intent.

[1726] 4. Virtual knowledge agent generation: Virtual knowledge agents with different thinking patterns are generated.

[1727] 5. Discussion and summary of results: The agents hold a discussion to reassure the nervous user and propose a strategy for displaying the product at eye level in the mid-price range.

[1728] 6. Displaying the results: The final proposal is displayed on the tablet, providing the user with specific advice such as "price the new product in the middle of the market and display it at eye level."

[1729] Prompt Sentence Examples

[1730] Example: Prompt text that the user types into the tablet

[1731] How to display new products

[1732] As described above, this invention recognizes the emotional state of the user and responds in real time, and virtual knowledge agents adaptively discuss the situation, thereby providing an optimal sales strategy for a physical store.

[1733] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1734] Step 1:

[1735] A user enters an assignment into a tablet

[1736] The user uses the tablet's user interface to input the problem they want to solve (e.g., "How to display a new product") in text format. The input problem text is saved on the device.

[1737] Input: User input text (issue)

[1738] Output: The assignment text is saved to your device

[1739] Step 2:

[1740] The device recognizes the user's emotions

[1741] While the user is entering the task, the device's camera captures the user's facial expressions and voice in real time, and an emotion recognition engine (EmotionRecognizer) is used to perform emotion analysis.

[1742] Input: User's facial expressions and voice data

[1743] Output: Emotional state data (e.g., tension, anxiety)

[1744] Step 3:

[1745] The device sends the task and emotion data to the server.

[1746] The task text and emotional state data are transmitted from the terminal to the server.

[1747] Input: Task text, emotional state data

[1748] Output: Data is sent to the server

[1749] Step 4:

[1750] The server parses the assignment text

[1751] The server uses natural language processing tools (NaiveBayesClassifier) ​​to analyze the submitted assignment text and extract key keywords and intent.

[1752] Input: Assignment text

[1753] Output: Analyzed data (keywords, task intent)

[1754] Step 5:

[1755] The server generates a virtual knowledge agent.

[1756] The server generates multiple virtual knowledge agents with different thinking patterns (e.g., sales maximization, profit maximization, cost minimization). Each agent analyzes the problem based on its own thinking pattern when it is generated.

[1757] Input: Analyzed data (keywords, problem intent)

[1758] Output: Multiple virtual knowledge agents

[1759] Step 6:

[1760] The server initiates a discussion with the virtual knowledge agent.

[1761] The virtual knowledge agents discuss issues from their own perspectives, taking into account their emotional states. During the discussion, emotional data is provided in real time, allowing the agents to adaptively adjust the content of the discussion based on this data.

[1762] Input: Multiple virtual knowledge agents, emotional state data

[1763] Output: Discussion results

[1764] Step 7:

[1765] The server aggregates the results of the discussions

[1766] The server aggregates the results of discussions among the virtual knowledge agents and derives an optimal conclusion, taking into account emotional data to arrive at a conclusion that is easily accepted by the user.

[1767] Input: Discussion results of virtual knowledge agents, emotional state data

[1768] Output: Aggregated optimal conclusion

[1769] Step 8:

[1770] The server sends the aggregated discussion results to the terminal.

[1771] The server sends the aggregated discussion results to the device, where users can review them and implement the proposed solutions.

[1772] Input: Aggregated discussion results

[1773] Output: The discussion results are sent to the terminal.

[1774] Step 9:

[1775] The device displays the best decision for the user.

[1776] The terminal displays the discussion results received from the server on the user interface. The user can review the proposed solutions (e.g., "The new product should be priced in the middle of the market and displayed at eye level") and use them to determine a sales strategy.

[1777] Input: Discussion results

[1778] Output: What is displayed to the user

[1779] 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.

[1780] 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.

[1781] 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.

[1782] 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.

[1783] 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.

[1784] 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.

[1785] 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).

[1786] 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.

[1787] 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."

[1788] 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.

[1789] 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).

[1790] 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.

[1791] 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.

[1792] 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.

[1793] 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.

[1794] 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.

[1795] 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.

[1796] 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.

[1797] 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.

[1798] 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.

[1799] 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.

[1800] The following is further disclosed regarding the above embodiment.

[1801] (Claim 1)

[1802] a user interface means for inputting tasks;

[1803] natural language processing means for analyzing the input problem;

[1804] A means for generating a plurality of virtual knowledge agents having different thought patterns;

[1805] means for initiating discussions among the virtual knowledge agents and aggregating the results of the discussions;

[1806] means for outputting the discussion results to a user interface means;

[1807] A system including:

[1808] (Claim 2)

[1809] 2. The system according to claim 1, wherein the plurality of virtual knowledge agents have thought patterns of sales maximization, profit maximization, and cost minimization.

[1810] (Claim 3)

[1811] 2. The system according to claim 1, wherein the natural language processing means comprises means for performing text analysis and extracting main keywords and intentions.

[1812] "Example 1"

[1813] (Claim 1)

[1814] a user interface means for inputting tasks;

[1815] natural language processing means for analyzing the input problem;

[1816] A means for generating a plurality of virtual knowledge agents having different thought patterns;

[1817] a means for presenting a problem to the virtual knowledge agents and having them individually analyze the problem;

[1818] means for initiating discussions among the virtual knowledge agents and aggregating the results of the discussions;

[1819] means for outputting the discussion results to a user interface means;

[1820] A system including:

[1821] (Claim 2)

[1822] 2. The system according to claim 1, wherein the plurality of virtual knowledge agents have thought patterns of sales maximization, profit maximization, and cost minimization.

[1823] (Claim 3)

[1824] 2. The system according to claim 1, wherein the natural language processing means comprises means for performing text analysis and extracting main keywords and intent.

[1825] "Application Example 1"

[1826] (Claim 1)

[1827] a user interface means for inputting tasks;

[1828] natural language processing means for analyzing the input problem;

[1829] A means for generating a plurality of virtual knowledge agents having different thought patterns;

[1830] means for initiating discussions among the virtual knowledge agents and aggregating the results of the discussions;

[1831] means for outputting the discussion results to a user interface means;

[1832] A means for recommending content based on user-inputted issues;

[1833] means for said virtual knowledge agents to analyze and evaluate content from their different perspectives;

[1834] A system including:

[1835] (Claim 2)

[1836] 10. The system of claim 1, wherein the plurality of virtual knowledge agents have sentiment analysis, trend following, and data-driven thinking patterns.

[1837] (Claim 3)

[1838] 2. The system according to claim 1, wherein the natural language processing means comprises means for performing text analysis and extracting main keywords and intentions.

[1839] "Example 2: Combining Emotion Engines"

[1840] (Claim 1)

[1841] a user interface means for inputting tasks;

[1842] natural language processing means for analyzing the input task and emotion data;

[1843] A means for generating a plurality of virtual knowledge agents having different thought patterns;

[1844] a means for initiating a discussion among the virtual knowledge agents, monitoring the emotional state of the user in real time during the discussion, and aggregating the results of the discussion;

[1845] means for outputting the discussion results to a user interface means;

[1846] A system including:

[1847] (Claim 2)

[1848] 2. The system according to claim 1, wherein the plurality of virtual knowledge agents have thought patterns of sales maximization, profit maximization, and cost minimization.

[1849] (Claim 3)

[1850] 2. The system according to claim 1, wherein the natural language processing means comprises means for performing text analysis and extracting main keywords and intentions.

[1851] "Application example 2 when combining emotion engines"

[1852] (Claim 1)

[1853] a user interface means for inputting tasks;

[1854] natural language processing means for analyzing the input problem;

[1855] A means for generating a plurality of virtual knowledge agents having different thought patterns;

[1856] means for initiating discussions among the virtual knowledge agents and aggregating the results of the discussions;

[1857] means for outputting the discussion results to a user interface means;

[1858] emotion recognition means for recognizing an emotional state of a user;

[1859] means for adaptively adjusting the virtual knowledge agent's arguments based on said emotional state;

[1860] A system including:

[1861] (Claim 2)

[1862] 2. The system according to claim 1, wherein the plurality of virtual knowledge agents have thought patterns of sales maximization, profit maximization, and cost minimization.

[1863] (Claim 3)

[1864] 2. The system according to claim 1, wherein the natural language processing means comprises means for performing text analysis and extracting main keywords and intentions. [Explanation of symbols]

[1865] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. a user interface means for inputting tasks; natural language processing means for analyzing the input problem; A means for generating a plurality of virtual knowledge agents having different thought patterns; means for initiating discussions among the virtual knowledge agents and aggregating the results of the discussions; means for outputting the discussion results to a user interface means; A system including:

2. 2. The system according to claim 1, wherein said plurality of virtual knowledge agents have thought patterns of sales maximization, profit maximization, and cost minimization.

3. 2. The system according to claim 1, wherein the natural language processing means comprises means for performing text analysis and extracting main keywords and intentions.

Images