System

Abstract

To provide a system.SOLUTION: The system includes: means for an AI to learn public minutes as teacher data; means for the learned AI to qualitatively evaluate the work of an assembly member; means for creating the result of the evaluation of the assembly member as a "report card of the assembly member; and outputting the result of the evaluation in a viewable manner.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] Currently, the criteria for voters to decide who to vote for in elections are unclear, and they often vote based on superficial information such as the candidate's photo and name recognition. However, voters should actually vote based on the candidate's specific work, such as their policies and activities in parliament. However, such information is generally not publicly available, or even if it is publicly available, there is no way to properly evaluate it. [Means for solving the problem]

[0005] This invention uses publicly available Diet records and other data as training data for AI to learn and quantitatively and qualitatively evaluate the performance of lawmakers. The evaluation results are compiled and made public as "lawmaker report cards," providing voters with clear criteria for deciding who to vote for. This allows voters to evaluate the specific performance of candidates before casting their vote. [Brief explanation of the drawings]

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

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

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

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

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

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

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

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

[0014] [First embodiment]

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

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

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

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

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

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

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

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

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

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

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

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

[0027] "Example 1"

[0028] One embodiment of the present invention is a system in which AI learns from publicly available Diet minutes and other training data. Specifically, the system acquires text data from Diet minutes and analyzes it using natural language processing technology. The analysis extracts information such as the content and frequency of Diet members' remarks, attendance rate, etc., and uses this information to quantitatively and qualitatively evaluate the performance of Diet members.

[0029] "Example 2"

[0030] The evaluation results are compiled as "report cards for legislators" and made public. Specifically, the evaluation results for each legislator are stored in a database, and an HTML report card is generated based on the results. The generated report card is made public on a website so that voters can view it.

[0031] "Example 3"

[0032] Another embodiment of the present invention is a system that uses the evaluation results as reference information when voters decide who to vote for. Specifically, the system refers to each legislator's report card on a website and decides who to vote for based on the evaluation results. For example, it is possible to give priority to legislators with high attendance rates or high evaluations of the content of their statements.

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

[0034] "Example 1"

[0035] Step 1: Obtain text data of the Diet minutes from the Internet.

[0036] Step 2: The acquired text data is analyzed using natural language processing technology. This analysis extracts information such as the content and frequency of the legislator's statements and attendance rate.

[0037] Step 3: Based on the extracted information, the AI ​​quantitatively and qualitatively evaluates the performance of the lawmakers.

[0038] "Example 2"

[0039] Step 1: Save the AI ​​evaluation results in a database.

[0040] Step 2: Generate an HTML report card based on the saved evaluation results.

[0041] Step 3: Publish the generated report card on a website so that voters can view it.

[0042] "Example 3"

[0043] Step 1: Voters view each representative's report card on the website.

[0044] Step 2: Voters decide who to vote for based on the evaluation results of the report card they have referenced. For example, they can prioritize lawmakers with high attendance rates or high evaluations of the content of their speeches.

[0045] Example 1

[0046] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0047] Conventional evaluation systems for legislators often require manual data collection and analysis to quantitatively and qualitatively evaluate legislators' performance, which is time-consuming and labor-intensive. Furthermore, evaluation results are not centrally managed and are not provided in an easily accessible format for users, making it difficult to use the evaluation results to decide who to vote for.

[0048] 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.

[0049] In this invention, the server includes: means for an artificial intelligence to learn using published minutes and the like as training data; means for the trained artificial intelligence to quantitatively and qualitatively evaluate the performance of legislators; means for creating and publishing the evaluation results as "legislator report cards"; means for acquiring text data of minutes via the Internet; means for preprocessing the acquired text data; means for analyzing the preprocessed data using natural language processing technology; means for evaluating the performance of legislators based on the analysis results; and means for outputting the evaluation results in a format that can be viewed by users. This makes it possible to efficiently and accurately evaluate the performance of legislators and provide the evaluation results in a format that users can easily refer to.

[0050] "Published minutes" are records of meetings that are made public by a government or public institution.

[0051] "Training data" is a dataset used for learning artificial intelligence.

[0052] "Artificial intelligence" is a technology that allows computer systems to behave in a way that mimics human intelligence.

[0053] "Quantitative evaluation" is a method of evaluating an object using numbers and statistics.

[0054] "Qualitative evaluation" is a method of evaluating an object by taking into account its qualitative aspects.

[0055] A "legislator report card" is a report summarizing the results of an evaluation of a member of parliament's work.

[0056] The "Internet" is a global communications network that interconnects computer networks around the world.

[0057] "Text data" is data that includes character information.

[0058] "Preprocessing" is the process of formatting and processing data before data analysis.

[0059] "Natural language processing technology" is a technology that uses computers to understand, analyze, and generate human language.

[0060] "Analysis" is the process of examining and evaluating data in detail.

[0061] "Evaluation results" are the results of evaluation obtained based on analysis.

[0062] "User" means a person or organization that uses this system.

[0063] "Format" is the specific way in which data or information is presented.

[0064] MODE FOR CARRYING OUT THE INVENTION

[0065] This invention is a system in which an artificial intelligence learns from publicly available minutes and other training data, and quantitatively and qualitatively evaluates the performance of legislators. A specific embodiment of this system will be described below.

[0066] 1. Program Generation

[0067] The server uses publicly available data such as meeting minutes as training data to build a generative AI model. This AI model is designed to analyze the text data in the minutes using natural language processing techniques. Specific software used includes Python's NLTK library and spaCy.

[0068] 2. Data Acquisition and Preprocessing

[0069] The terminal acquires the text data of the minutes via the Internet. The acquired data is sent to the server, which then performs the following preprocessing.

[0070] Text cleaning: Remove unnecessary symbols and spaces.

[0071] Tokenization: Splitting text into words and sentences.

[0072] Stop word removal: Remove meaningless words (e.g. "no", "wa", etc.).

[0073] 3. Data Analysis

[0074] The server analyzes the preprocessed data using natural language processing technology. Specifically, it performs the following analysis:

[0075] Speech analysis: Analyze the topic and sentiment of the speech.

[0076] Calculating speech frequency: Count the number of times each member of parliament speaks.

[0077] Calculating attendance rate: Extract the attendance rate of each member from the minutes.

[0078] 4. Generating ratings

[0079] The server evaluates the performance of the legislators based on the analysis results. The evaluation is performed as follows:

[0080] Quantitative evaluation: Quantify speaking frequency and attendance rate.

[0081] Qualitative evaluation: Evaluating the quality of what was said.

[0082] 5. Outputting the results

[0083] The server outputs the evaluation results in a format that can be viewed by the user, such as a web interface or a PDF report.

[0084] Specific examples

[0085] Specific use cases

[0086] If a user wants to evaluate the performance of a particular legislator, they can input a prompt like the following into the generative AI model:

[0087] Example prompt sentence:

[0088] "Based on the 2023 Diet minutes, please evaluate the content, frequency, and attendance rate of Congressman A."

[0089] When this prompt is entered, the following specific processing is performed:

[0090] 1. Data Acquisition:

[0091] The device will download the 2023 Diet minutes from the government's public API.

[0092] 2. Data preprocessing:

[0093] The server removes unnecessary symbols and spaces from the downloaded text data and performs tokenization.

[0094] The server removes stop words and converts the data into a format suitable for analysis.

[0095] 3. Data Analysis:

[0096] The server uses natural language processing technology to analyze the content of Congressman A's statements. For example, it detects that Congressman A frequently speaks about "economic policy."

[0097] The server calculates the frequency of Congressman A's statements and finds that the average number of times per month is 20.

[0098] The server calculates the attendance rate of Council Member A and extracts that it is 95%.

[0099] 4. Generate ratings:

[0100] The server quantitatively evaluates the performance of Councillor A based on frequency of speech and attendance rate.

[0101] The server qualitatively evaluates the performance of Congressman A based on the quality of his remarks.

[0102] 5. Result output:

[0103] The server generates the evaluation results as a PDF report that can be downloaded by the user.

[0104] In this way, the user can evaluate Congressman A's performance in detail.

[0105] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0106] Program processing flow

[0107] Step 1: Get the data

[0108] Input: The user enters a prompt to rate the performance of a particular legislator.

[0109] Operation: The device downloads text data of meeting minutes for a specified period from the government's public API via the Internet.

[0110] Output: Downloaded text data of the minutes.

[0111] Step 2: Preprocessing the data

[0112] Input: Text data of downloaded minutes.

[0113] Operation: The server performs the following preprocessing.

[0114] Text cleaning: Remove unnecessary symbols and spaces.

[0115] Tokenization: Splitting text into words and sentences.

[0116] Stop word removal: Remove meaningless words (e.g. "no", "wa", etc.).

[0117] Output: Preprocessed text data.

[0118] Step 3: Analyze the data

[0119] Input: Preprocessed text data.

[0120] Operation: The server performs the following analysis using natural language processing technology.

[0121] Speech analysis: Analyze the topic and sentiment of the speech.

[0122] Calculating speech frequency: Count the number of times each member of parliament speaks.

[0123] Calculating attendance rate: Extract the attendance rate of each member from the minutes.

[0124] Output: Analysis results (content of comments, frequency of comments, attendance rate).

[0125] Step 4: Generate ratings

[0126] Input: Analysis results (content of comments, frequency of comments, attendance rate).

[0127] How it works: The server evaluates the performance of the legislators based on the analysis results.

[0128] Quantitative evaluation: Quantify speaking frequency and attendance rate.

[0129] Qualitative evaluation: Evaluating the quality of what was said.

[0130] Output: Evaluation results (quantitative evaluation, qualitative evaluation).

[0131] Step 5: Output the results

[0132] Input: Evaluation results (quantitative evaluation, qualitative evaluation).

[0133] Behavior: The server outputs the evaluation results in a format that can be viewed by the user, such as a web interface or a PDF report.

[0134] Output: User-visible evaluation results.

[0135] Specific examples

[0136] Specific use cases

[0137] If a user wants to evaluate the performance of a particular legislator, they can input a prompt like the following into the generative AI model:

[0138] Example prompt sentence:

[0139] "Based on the 2023 Diet minutes, please evaluate the content, frequency, and attendance rate of Congressman A."

[0140] When this prompt is entered, the following specific processing is performed:

[0141] 1. Data Acquisition:

[0142] The device will download the 2023 Diet minutes from the government's public API.

[0143] 2. Data preprocessing:

[0144] The server removes unnecessary symbols and spaces from the downloaded text data and performs tokenization.

[0145] The server removes stop words and converts the data into a format suitable for analysis.

[0146] 3. Data Analysis:

[0147] The server uses natural language processing technology to analyze the content of Congressman A's statements. For example, it detects that Congressman A frequently speaks about "economic policy."

[0148] The server calculates the frequency of Congressman A's statements and finds that the average number of times per month is 20.

[0149] The server calculates the attendance rate of Council Member A and extracts that it is 95%.

[0150] 4. Generate ratings:

[0151] The server quantitatively evaluates the performance of Councillor A based on frequency of speech and attendance rate.

[0152] The server qualitatively evaluates the performance of Congressman A based on the quality of his remarks.

[0153] 5. Result output:

[0154] The server generates the evaluation results as a PDF report that can be downloaded by the user.

[0155] In this way, the user can evaluate Congressman A's performance in detail.

[0156] (Application example 1)

[0157] Next, a description will be given of Application Example 1 of Embodiment Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0158] In traditional conferences and meetings, it was difficult to quantitatively and qualitatively evaluate participants' comments, frequency of comments, attendance rates, etc., resulting in problems that reduced the transparency and efficiency of the meetings. Furthermore, there was a lack of means to objectively evaluate participants' performance, making it impossible to use this information as reference for decision-making. This could result in a decline in the quality of meetings and a negative impact on the performance of the entire organization.

[0159] 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.

[0160] In this invention, the server includes a means for AI to learn using publicly available minutes and the like as training data, a means for the trained AI to quantitatively and qualitatively evaluate the performance of participants, a means for creating and publishing the evaluation results as a "participant report card," and a means for improving the transparency and efficiency of meetings. This makes it possible to objectively evaluate the content and frequency of participants' comments, attendance rates, etc., and improve the transparency and efficiency of meetings.

[0161] A "published minutes" is a document that records the contents of a conference or meeting and is available to the public.

[0162] "Training data" is a dataset used to train a machine learning model.

[0163] "AI" is an abbreviation for artificial intelligence, a system that mimics human intellectual tasks using technologies such as machine learning and natural language processing.

[0164] A "participant" is someone who attends a conference or meeting and expresses their opinions or comments.

[0165] "Quantitative evaluation" is a method of evaluation using numerical and statistical data.

[0166] "Qualitative evaluation" is a method of evaluation that takes into account qualitative aspects.

[0167] A "participant report card" is a document of evaluation results created based on the content and frequency of participants' comments, attendance rate, etc.

[0168] "Transparency" means that information is not hidden but is made public and available for anyone to see.

[0169] "Efficiency" is the state in which the time and effort required to achieve a goal is minimized.

[0170] A system for carrying out this invention is configured as follows: The server includes means for AI to learn using publicly available minutes and the like as training data, means for the trained AI to quantitatively and qualitatively evaluate the performance of participants, means for creating and publishing the evaluation results as "participant report cards," and means for improving the transparency and efficiency of meetings.

[0171] The server analyzes the minutes text using Python and the natural language processing library spaCy. Specifically, the server performs the following steps:

[0172] First, the server reads the minutes text from a file. Next, it analyzes the read text using spaCy to extract the content of each speaker's comments. The extracted content is categorized by speaker and the frequency of each comment is calculated. This allows for a quantitative evaluation of the content and frequency of each participant's comments.

[0173] The evaluation results are compiled as a "participant report card" and made publicly available for users to refer to, allowing them to use the information to improve the transparency and efficiency of their meetings.

[0174] As a concrete example, let's say we have the following meeting minutes:

[0175] Tanaka: Let's discuss a new project today.

[0176] Suzuki: Before that, let's check the minutes of the last meeting.

[0177] Tanaka: Understood. I will check the minutes.

[0178] An analysis of the minutes reveals the following results:

[0179] Tanaka: 2 comments

[0180] Suzuki: 1 statement

[0181] An example of a prompt sentence to input to the generative AI model is as follows:

[0182] Analyze the following meeting minutes and calculate the frequency of each speaker's comments.

[0183] Tanaka: Let's discuss a new project today.

[0184] Suzuki: Before that, let's check the minutes of the last meeting.

[0185] Tanaka: Understood. I will check the minutes.

[0186] In this way, the server can analyze the content and frequency of participants' comments and publish the evaluation results as a "participant report card," thereby improving the transparency and efficiency of meetings.

[0187] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0188] Step 1:

[0189] The server reads the published minutes text from a file. The input is the minutes text file, and the output is text data. This text data serves as the basis for subsequent natural language processing.

[0190] Step 2:

[0191] The server analyzes the loaded text data using spaCy, a natural language processing library. The input is the text data obtained in step 1, and the output is a parsed document object. This document object contains information such as sentence and word tokenization, part-of-speech tagging, and dependency analysis.

[0192] Step 3:

[0193] The server extracts the content of each speaker's comments from the analyzed document object. The input is the document object obtained in step 2, and the output is a list of comments categorized by speaker. Specifically, the speaker and the content of the comments are separated based on the ":" in the sentence, and the list is organized by speaker.

[0194] Step 4:

[0195] The server calculates the frequency of each speaker's comments. The input is the list of comments obtained in step 3, and the output is a dictionary of the frequency of each speaker's comments. Specifically, the server counts the number of comments made by each speaker and saves them in dictionary format.

[0196] Step 5:

[0197] The server creates a "participant report card" based on the dictionary of comment frequency. The input is the dictionary of comment frequency obtained in step 4, and the output is the evaluation results as a "participant report card." Specifically, the evaluation of each participant is quantitatively described based on the comment frequency and saved in document format.

[0198] Step 6:

[0199] The server publishes the created "Participant Report Cards" in a form that users can refer to. The input is the evaluation results obtained in step 5, and the output is a web page or dashboard that users can access. Specifically, the evaluation results are converted into HTML format and published on the web server.

[0200] In this way, the server analyzes the minutes text, evaluates the content and frequency of participants' comments, and publishes the results as a "participant report card," thereby improving the transparency and efficiency of meetings.

[0201] Example 2

[0202] Next, a description will be given of Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0203] In the conventional system for evaluating legislators, the evaluation results are not managed centrally, and the evaluations lack transparency and reliability. In addition, the evaluation results are not properly disclosed to the general public, which creates the problem that voters cannot use them as reference information when deciding who to vote for.

[0204] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes a means for an artificial intelligence to learn using published minutes and the like as training data, a means for the trained artificial intelligence to quantitatively and qualitatively evaluate the performance of the legislators, a means for saving the evaluation results in a database, a means for generating a report card in HTML format based on the saved evaluation results, and a means for publishing the generated report card on a website. This enables centralized management of evaluation results and ensures transparency, allowing voters to decide who to vote for based on reliable information.

[0205] "Published minutes, etc." are documents and data that record the content of statements and discussions made in official forums such as parliaments and conferences and are made publicly available.

[0206] "Artificial intelligence" is a technology that enables computer systems to mimic human intelligence and perform intellectual tasks such as learning, reasoning, and recognition.

[0207] The "means of learning" refers to the process by which AI analyzes data, such as publicly available minutes, to find patterns and rules.

[0208] "Quantitative and qualitative evaluation methods" refers to the process by which artificial intelligence evaluates the performance of legislators using numerical data (quantitative) and qualitative data (qualitative).

[0209] "Evaluation results" are data obtained as a result of AI evaluating the performance of lawmakers, and include the activities, statements, attendance rate, etc. of lawmakers.

[0210] A "database" is a digital information management system for efficiently storing, managing, and searching evaluation results.

[0211] An "HTML report card" is a web page document created based on the evaluation results to visually display the evaluation of a member of parliament.

[0212] "Publication on a website" is the process by which the generated HTML report card is made accessible to the public on the Internet.

[0213] "Users" are ordinary voters and users who access the website and view the report cards of their representatives.

[0214] This invention relates to a system for evaluating the performance of assembly members and publishing the evaluation results as a “asset member report card.” A specific embodiment of this system will be described below.

[0215] 1. Data collection and storage

[0216] The server collects publicly available minutes and other training data, including official records of parliaments and conferences. The collected data is stored in a MySQL (registered trademark) database. Specifically, the evaluation results are inserted into the database using INSERT statements.

[0217] 2. Data processing

[0218] The server retrieves the evaluation results from the MySQL database. Specifically, it extracts the required data using a SELECT statement. The retrieved data is converted into a data frame using the Python Pandas library and formatted as needed.

[0219] 3. Report Card Generation

[0220] The server uses the Jinja2 template engine to generate an HTML report card based on the processed data. Specifically, it generates the HTML by embedding the data in a template file. The generated HTML file is saved in the public directory of the web server.

[0221] 4. Publication of report cards

[0222] The server uses the Flask framework to build a web server. An endpoint for displaying report cards is set up and made accessible to users. A user accesses the website and views the report card of a specific legislator. Specifically, by entering the legislator's name in the search bar and clicking the search button, the report card of the corresponding legislator is displayed.

[0223] Specific examples

[0224] For example, consider a case where a user accesses a website and wants to view the report card of a particular politician. The user enters the politician's name in the website's search bar and clicks the search button. The server retrieves the evaluation results of the relevant politician from the database and generates and displays the report card in HTML format.

[0225] Prompt Sentence Examples

[0226] "Write a program that generates an HTML report card based on the evaluation results of a specific council member and publishes it on a website. The evaluation results are stored in a MySQL database, and you will use Python's Pandas library and the Jinja2 template engine to process the data and generate the report card. You will then publish the report card on the website using the Flask framework."

[0227] The above is a specific embodiment for carrying out the present invention.

[0228] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0229] Step 1:

[0230] The server collects publicly available minutes and other data as training data. Specifically, it obtains official records of parliaments and meetings through APIs and manual input forms. Input data includes what council members said, attendance rates, and activities. This data is stored in a MySQL database. Specifically, it inserts the evaluation results into the database using INSERT statements. The output is the evaluation results stored in the database.

[0231] Step 2:

[0232] The server retrieves the evaluation results from the MySQL database. Specifically, it uses a SELECT statement to extract the necessary data. The input is the evaluation results stored in the database. The retrieved data is converted into a data frame using the Python Pandas library. The converted data frame is then formatted as needed. For example, the evaluation scores are sorted in ascending order. The output is a formatted data frame.

[0233] Step 3:

[0234] The server uses the Jinja2 template engine to generate an HTML report card based on the processed data. The input is a formatted data frame. Specifically, HTML is generated by embedding data in a template file. The generated HTML file is saved in the public directory of the web server. The output is the generated HTML report card.

[0235] Step 4:

[0236] The server uses the Flask framework to build a web server. An endpoint for displaying report cards is set up. The input is the generated report card in HTML format. A user accesses the website and wants to view the report card of a specific legislator. Specifically, they enter the legislator's name in the search bar and click the search button. The server receives the user's request and displays the report card of the relevant legislator. The output is the report card displayed in the user's browser.

[0237] (Application example 2)

[0238] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0239] In conventional evaluation systems, it was difficult for users to easily refer to evaluation results based on the target's activities, comments, attendance rate, etc., and use them as a reference for decision-making. In addition, there was a lack of a way to save the evaluation results in a database and publish them in HTML format, making it difficult for users to visually check the evaluation results and post reviews.

[0240] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes a means for the artificial intelligence to learn using published minutes and the like as training data, a means for the learned artificial intelligence to quantitatively and qualitatively evaluate the target's performance, a means for creating and publishing the evaluation results as a "report card," a means for saving the evaluation results in a database and generating the report card in HTML format, and a means for users to refer to the evaluation results and post reviews. This allows users to visually check the evaluation results and use them as a reference for decision-making.

[0241] "Published minutes, etc." refers to minutes and related documents that are made publicly accessible.

[0242] "Artificial intelligence" refers to systems that mimic human intelligence through machine learning and data analysis.

[0243] "Means of learning" refers to the process by which artificial intelligence analyzes data based on publicly available minutes and other documents to find patterns and rules.

[0244] "Quantitative and qualitative evaluation methods" refers to the process of evaluating the functioning of an object using numerical and qualitative data.

[0245] A "report card" refers to a report-style document summarizing evaluation results.

[0246] "Disclosure means" refers to the process of making evaluation results publicly accessible.

[0247] "Means for storing in a database" refers to the process of storing the evaluation results in a structured format in a database.

[0248] "Means for generating an HTML report card" refers to the process of generating the evaluation results as an HTML document.

[0249] "Means for users to view rating results and post reviews" refers to an interface that allows users to view rating results and add their own opinions and ratings.

[0250] To implement the present invention, the following system configuration is required.

[0251] System Configuration

[0252] 1. Server

[0253] Hardware: A server with a powerful processor, sufficient memory, and storage.

[0254] Software: Python, Flask, SQLite, generative AI models.

[0255] 2. Terminal

[0256] Hardware: Smartphone.

[0257] Software: Web browser or dedicated application.

[0258] 3. Users

[0259] Role: View ratings and post reviews.

[0260] Program processing explanation

[0261] server

[0262] The server provides a means for the AI ​​to learn from publicly available meeting minutes and other data as training data. Specifically, the generative AI model is used to analyze the meeting minutes data and quantitatively and qualitatively evaluate the target's performance. The evaluation results are stored in a database and generated as a report card in HTML format. This allows users to visually check the evaluation results and use them as a reference for decision-making.

[0263] Terminal

[0264] The terminal provides an interface that allows users to view the evaluation results and post reviews. Using their smartphones, users can view the HTML report card provided by the server and add their own opinions and evaluations.

[0265] Specific examples

[0266] For example, when a user opens the app on their smartphone, they are presented with a list of nearby stores' ratings. Each store's rating includes reviews and star ratings posted by other users. Users can post new reviews, which are then sent to the server and stored in a database.

[0267] Prompt Sentence Examples

[0268] "Create a smartphone app that displays the ratings of nearby businesses and allows users to post reviews. The ratings will be stored in a database and displayed in HTML format."

[0269] The above is a specific embodiment for carrying out the present invention.

[0270] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0271] Step 1:

[0272] The server collects published minutes and stores them in a database. The input is the published minutes and other data, and the output is the raw data stored in the database. Specifically, the minutes data is obtained using web scraping and APIs and stored in an SQLite database.

[0273] Step 2:

[0274] The server uses a generative AI model to analyze the minutes data stored in the database and quantitatively and qualitatively evaluate the subject's performance. The input is the minutes data stored in the database, and the output is the evaluation results. Specifically, it uses natural language processing technology to analyze the minutes data and evaluate each subject's activities, comments, attendance rate, etc.

[0275] Step 3:

[0276] The server saves the evaluation results in a database and generates a report card in HTML format. The input is the evaluation results, and the output is an HTML report card. Specifically, it generates an HTML template based on the evaluation results and saves it in the database.

[0277] Step 4:

[0278] The server publishes the report card in HTML format through a web server so that users can view the evaluation results. The input is the report card in HTML format, and the output is the report card displayed in the user's browser. Specifically, a web server is built using Flask so that users can access it.

[0279] Step 5:

[0280] Users access the server using a smartphone browser or a dedicated app to view the evaluation results. The input is the user's access request, and the output is a display of the evaluation results. Specifically, the user opens a browser and accesses the server's URL, and the report card is displayed.

[0281] Step 6:

[0282] Users post reviews based on the evaluation results. The input is the user's review content, and the output is the review sent to the server. Specifically, the user enters the review in the form and presses the submit button, which sends the review to the server.

[0283] Step 7:

[0284] The server stores the reviews sent by users in a database and makes them available for other users to refer to. The input is the user's review content, and the output is the review saved in the database. Specifically, the review content is stored in the database and reflected in an HTML report card.

[0285] The above are the specific processing steps for carrying out the present invention.

[0286] Example 3

[0287] Next, a description will be given of a third embodiment of the third embodiment. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0288] In conventional elections, it is difficult for voters to fully understand the activities and evaluations of their representatives, resulting in a lack of information to decide who to vote for. In addition, there is a lack of means to effectively utilize publicly available parliamentary records and quantitatively and qualitatively evaluate the activities of representatives, making it difficult for voters to decide who to vote for based on reliable information.

[0289] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[0290] In this invention, the server includes means for the artificial intelligence to learn using publicly available parliamentary records and the like as training data, means for the trained artificial intelligence to quantitatively and qualitatively evaluate the activities of legislators, means for creating and publishing the evaluation results as "legislator report cards," means for displaying the evaluation results on a website, and means for users to refer to the evaluation results on the website and decide who to vote for. This enables voters to evaluate the activities of legislators based on reliable information and decide who to vote for.

[0291] "Public parliamentary records" are information such as minutes, attendance records, and statements made that are officially recorded in parliament and made available to the public.

[0292] "Artificial intelligence" is a system that uses machine learning and data analysis techniques to learn patterns from data and make predictions and classifications.

[0293] "The activities of a member of parliament" are the actions and duties of a member of parliament, such as attending parliament, speaking, proposing bills, and voting.

[0294] "Quantitative and qualitative evaluation" means conducting both quantitative evaluation using numerical data and qualitative evaluation that evaluates content and quality.

[0295] A "legislator report card" is a report summarizing the activities and evaluation results of a member of parliament, and is a resource that helps voters understand the performance of their members.

[0296] A "website" is an online page published on the Internet that users can access and view information.

[0297] A "user" is a voter who uses the system to obtain information and decide how to vote.

[0298] "Training data" is a dataset consisting of known input and output pairs that is used to train an artificial intelligence.

[0299] "Learning" is the process by which artificial intelligence finds patterns and rules based on training data, improving the accuracy of predictions and classifications.

[0300] The "evaluation results" are the results of the AI ​​analyzing the activities of lawmakers and evaluating them quantitatively and qualitatively.

[0301] This invention is a system that provides evaluation results that voters can use as a reference when deciding who to vote for. Specifically, a server collects publicly available parliamentary records, evaluates the activities of legislators using artificial intelligence, and displays the evaluation results on a website. Users can refer to the evaluation results on the website and decide how to vote appropriately.

[0302] Hardware and software used

[0303] 1. Server

[0304] The server collects data, processes the data, generates evaluation results, and displays them on the website.

[0305] Software used includes Python, BeautifulSoup, Pandas, Scikit-learn, Flask or Django.

[0306] 2. Terminal

[0307] The terminal is used by the user to access the website and view the evaluation results.

[0308] The hardware used includes computers, smartphones, tablets, etc.

[0309] The software used includes a web browser (such as GOOGLE CHROME (registered trademark), Mozilla Firefox, or Safari).

[0310] 3. Users

[0311] Users access the website as voters and decide who to vote for based on the evaluation results.

[0312] Program processing

[0313] 1. Data Collection

[0314] The server collects data such as each member's attendance rate and speech content from the parliament's official website and publicly available minutes.

[0315] Example: The server uses Python's BeautifulSoup library to scrape HTML data from the official parliamentary website, for example, parsing the page listing the attendance rates of members of parliament and extracting the required data.

[0316] 2. Data processing

[0317] The server organizes the collected data and creates a report card for each member of parliament, including attendance rate, evaluation of speech content, number of bills proposed, and more.

[0318] Example: The server uses the Pandas library to convert the collected data into a data frame. Then, it extracts the necessary information for each member of parliament and formats it into a report card format. For example, it extracts the attendance column and stores it in a new data frame along with each member's name.

[0319] 3. Generating evaluation results

[0320] The server generates evaluation results for each member of parliament, which are scored based on their attendance rate and the content of their comments.

[0321] Example: The server uses the Scikit-learn library to build a model for scoring the assessment results, for example a linear regression model that assigns higher scores to students with higher attendance rates.

[0322] 4. Display on the Website

[0323] The server displays the generated evaluation results on a website, and users can access the website to view the report cards of each member of parliament.

[0324] Example: The server uses a web framework such as Flask or Django to create a web page that displays the evaluation results. For example, it generates an HTML page that displays the names of each legislator and their evaluation scores in a table.

[0325] 5. User voting decision

[0326] Users can refer to the report cards of each member of parliament on the website and decide who to vote for based on the evaluation results.

[0327] Example: A user accesses a website using a browser, checks the displayed ratings, and, for example, selects the legislator with the highest rating score and decides to vote for that legislator.

[0328] Prompt Sentence Examples

[0329] Examples of prompts to be input to a generative AI model include:

[0330] "In order to decide which representative to vote for in the next election, I would like to refer to a report card based on the attendance rate and comments of each representative. I would like to check the evaluation results on the website and choose the representative with the highest evaluation."

[0331] By using this prompt sentence, the generative AI model can understand what information the user is looking for and provide an appropriate evaluation result. The flow of the identification process in Example 3 will be described with reference to Figure 15.

[0332] Step 1:

[0333] Data collection

[0334] The server collects data such as each member's attendance rate and speech content from the parliament's official website and publicly available minutes.

[0335] Input: URL of the official parliament website or publicly available minutes.

[0336] What it does: The server uses Python's BeautifulSoup library to scrape HTML data from the official parliamentary website. For example, it parses the page listing the attendance rates of members of parliament and extracts the necessary data.

[0337] Output: Raw data such as attendance rate and speech content of each member of parliament.

[0338] Step 2:

[0339] Data Processing

[0340] The server organizes the collected data and creates a report card for each member of parliament, including attendance rate, evaluation of speech content, number of bills proposed, and more.

[0341] Input: The raw data collected in step 1.

[0342] What it does: The server uses the Pandas library to convert the collected data into a data frame. Then, it extracts the necessary information for each member of parliament and formats it into a report card format. For example, it extracts the attendance column and stores it in a new data frame along with each member's name.

[0343] Output: Report card data for each member of parliament.

[0344] Step 3:

[0345] Generating evaluation results

[0346] The server generates evaluation results for each member of parliament, which are scored based on their attendance rate and the content of their comments.

[0347] Input: Report card data created in step 2.

[0348] What it does: The server uses the Scikit-learn library to build a model for scoring the assessment results, for example, a linear regression model that assigns higher scores to students with higher attendance rates.

[0349] Output: Evaluation scores for each legislator.

[0350] Step 4:

[0351] Display on the website

[0352] The server displays the generated evaluation results on a website, and users can access the website to view the report cards of each member of parliament.

[0353] Input: The rating scores generated in step 3.

[0354] What it does: The server uses a web framework such as Flask or Django to create a web page that displays the evaluation results. For example, it generates an HTML page that displays the names of each legislator and their evaluation scores in a table.

[0355] Output: A web page that displays the evaluation results.

[0356] Step 5:

[0357] User voting decision

[0358] Users can refer to the report cards of each member of parliament on the website and decide who to vote for based on the evaluation results.

[0359] Input: The assessment results displayed on the website.

[0360] Specific operation: A user accesses a website using a browser, checks the displayed evaluation results, and, for example, selects the legislator with the highest evaluation score and decides to vote for that legislator.

[0361] Output: User's vote decision.

[0362] (Application example 3)

[0363] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0364] In conventional lawmaker evaluation systems, evaluation results are simply published on a website, and users lack sufficient interactivity and convenience when deciding who to vote for based on that information. Furthermore, there is a lack of means to provide information in physical stores, and there is no system in place to allow users to check evaluation results in real time and receive recommendations on the optimal candidate to vote for.

[0365] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[0366] In this invention, the server includes: means for AI to learn using published Diet minutes and the like as training data; means for the trained AI to quantitatively and qualitatively evaluate the performance of Diet members; means for creating and publishing the evaluation results as "Diet member report cards"; means for displaying information on smartphones using QR codes (registered trademark) in conjunction with interactive displays in physical stores; means for recommending optimal voting destinations to users based on the evaluation results; and means for updating the evaluation results and recommendation information in real time. This enables users to interactively check the evaluation results in physical stores and receive optimal voting destination recommendations in real time.

[0367] "Published Diet minutes" are documents that record the discussions and statements made in the Diet and are available to the public.

[0368] "Teacher data" is a dataset used for learning by a machine learning model, and is data that has been assigned a correct answer label.

[0369] "AI" is an abbreviation for artificial intelligence, a technology that allows computers to imitate human intelligence to learn and reason.

[0370] "The work of a member of parliament" is a general term for the activities, statements, attendance, etc. that members of parliament undertake in the Diet and other official duties.

[0371] "Quantitative and qualitative evaluation" means evaluation using both numerical data (quantitative) and qualitative data (qualitative).

[0372] A "Member of Parliament Report Card" is a report summarizing the evaluation results of a member of parliament's activities, comments, attendance rate, etc.

[0373] An "interactive display in a physical store" is a display device that can be directly operated by a user and is installed in a physical store.

[0374] A "QR code" is a type of two-dimensional barcode that can be read by devices such as smartphones.

[0375] A "smartphone" is a mobile phone that can connect to the Internet and use applications.

[0376] "Recommending the best voting option" means suggesting the most suitable voting option for the user.

[0377] "Real-time updates" means that data and information are updated immediately and always kept up to date.

[0378] The system for implementing this invention is configured as follows: First, the server has a means for the AI ​​to learn from publicly available Diet minutes and other training data. This AI is used to quantitatively and qualitatively evaluate the performance of Diet members. The evaluation results are compiled as "Diet member report cards" and made public.

[0379] The second step involves linking the system to interactive displays in physical stores and displaying information on smartphones using QR codes. Users can check the evaluation results of each legislator by scanning the QR code displayed on the display in the store with their smartphone.

[0380] Furthermore, the server has a means to recommend the best voting destination to the user based on the evaluation results. This recommendation is made in real time based on the evaluation result data. The evaluation results and recommendation information are always kept up to date.

[0381] The specific hardware used is a smartphone and an interactive display in a physical store. The software used is Flask (a Python web framework) and requests (a library for sending HTTP requests). Data processing involves obtaining the evaluation results of each legislator in JSON format and sorting them to recommend the optimal voting option to the user. Data calculation involves executing recommendation logic based on the evaluation results of attendance rate, speech content, and policy performance.

[0382] For example, when a user scans a QR code displayed on a display in a physical store, the evaluation results for each politician are displayed on their smartphone. Based on this information, the user can confirm who the application recommends for voting.

[0383] An example of a prompt is, "Please recommend the best candidate to vote for based on your evaluation of each member of parliament's attendance rate, speech content, and policy achievements."

[0384] In this way, users can interactively check the evaluation results in a physical store and receive recommendations for the optimal voting destination in real time.

[0385] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[0386] Step 1:

[0387] The server collects publicly available Diet minutes and other data as training data. It receives Diet minutes data as input and prepares it as training data for the AI ​​model. It processes the data by cleaning the text data and adding necessary labels. It outputs a dataset in a format that the AI ​​model can learn from.

[0388] Step 2:

[0389] The server trains the AI ​​model using the collected training data. The dataset generated in step 1 is used as input. As data calculation, the AI ​​model learns patterns to quantitatively and qualitatively evaluate the performance of legislators. The output is a trained AI model.

[0390] Step 3:

[0391] The server evaluates the performance of lawmakers using a trained AI model. As input, it receives the latest parliamentary minutes data and inputs this into the AI ​​model. The data is then calculated to evaluate the lawmakers' attendance rates, statements, policy achievements, etc. The output is the evaluation results for each lawmaker.

[0392] Step 4:

[0393] The server creates the evaluation results as a "report card for the legislators" and publishes them on a website. The evaluation results obtained in step 3 are used as input. The data is processed by formatting the evaluation results into an easy-to-read format. The output is a web page that can be viewed by users.

[0394] Step 5:

[0395] The user scans a QR code displayed on an interactive display in a physical store with their smartphone. As input, the smartphone reads the information in the QR code. Specifically, the smartphone's camera is used to scan the QR code. As output, the smartphone obtains the URL of the page displaying the evaluation results.

[0396] Step 6:

[0397] The terminal (smartphone) accesses the server using the URL obtained from the QR code and displays the evaluation results. The URL obtained from the QR code is used as input. As data calculation, the evaluation result data is obtained from the server and displayed on the smartphone screen. As output, a screen is provided where the user can check the evaluation results.

[0398] Step 7:

[0399] The server recommends the optimal voting destination to the user based on the evaluation results. The evaluation results viewed by the user are used as input. The data calculation involves sorting the evaluation results and calculating the optimal voting destination. The output shows the voting destination recommended to the user.

[0400] Step 8:

[0401] The server updates the evaluation results and recommendation information in real time. It receives new Diet minutes data and user feedback as input. It processes the data by retraining the AI ​​model with the latest data. It always provides the latest evaluation results and recommendation information as output.

[0402] 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.

[0403] "Example 1"

[0404] One embodiment of the present invention is a system that incorporates an emotion engine. In this system, AI learns from publicly available Diet records and other training data, and the trained AI quantitatively and qualitatively evaluates the performance of Diet members. The evaluation results are compiled and made public as a "Diet member report card." Furthermore, this system includes an emotion engine that recognizes the user's emotions.

[0405] "Example 2"

[0406] The emotion engine recognizes the user's emotions and reflects them in the evaluation results. Specifically, the system captures the user's reaction when they view the lawmaker's report card via camera and microphone, and analyzes their emotions from their facial expressions and tone of voice. For example, if a user feels happy, angry, or surprised when they see the lawmaker's evaluation results, that emotion is recognized by the emotion engine and reflected in the next evaluation results.

[0407] "Example 3"

[0408] The emotion engine also analyzes the user's facial expressions, voice, and text input to recognize their emotions. For example, if a user comments on social media in response to the evaluation results of a lawmaker, "This lawmaker is doing a great job," the text conveys the user's positive emotions. This information is also collected and analyzed by the emotion engine and reflected in the evaluation results.

[0409] The processing flow of each embodiment will be described below.

[0410] "Example 1"

[0411] Step 1: The AI ​​learns from publicly available parliamentary records and other training data.

[0412] Step 2: The trained AI evaluates the performance of lawmakers quantitatively and qualitatively.

[0413] Step 3: Compile the evaluation results into a "report card for the legislator."

[0414] Step 4: Publish the "Lawmakers' Report Cards."

[0415] Step 5: The emotion engine recognizes the user's emotion.

[0416] "Example 2"

[0417] Step 1: The user looks at the legislator's report card.

[0418] Step 2: The emotion engine captures the user's reaction via camera and microphone.

[0419] Step 3: The emotion engine analyzes the user's emotions from their facial expressions and tone of voice.

[0420] Step 4: Reflect the analyzed emotions when creating the next evaluation results.

[0421] "Example 3"

[0422] Step 1: Users comment on the results of the lawmaker's evaluation on social media.

[0423] Step 2: The sentiment engine analyzes the user's sentiment from the text.

[0424] Step 3: Reflect the analyzed information in the evaluation results.

[0425] Example 1

[0426] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0427] Conventional lawmaker evaluation systems mainly evaluated lawmakers based on quantitative data such as their speech content and attendance rate, and were unable to fully consider the emotions and qualitative aspects of their speech. Furthermore, when users looked at the evaluation results to decide who to vote for, they were not provided with a comprehensive evaluation that included emotional factors, resulting in a lack of information to support users' decision-making.

[0428] 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.

[0429] In this invention, the server includes means for acquiring published minutes data, means for preprocessing the acquired data, means for analyzing the preprocessed data using natural language processing technology, means for extracting information such as the content of remarks, frequency of remarks, and attendance rate from the analysis results, means for quantitatively and qualitatively evaluating the performance of the legislator based on the extracted information, means for creating and publishing the evaluation results as a "legislator report card," and an emotion engine that recognizes the user's emotions. This enables comprehensive evaluations that take into account not only the content of legislators' remarks and attendance rate, but also the emotions and qualitative aspects contained in the remarks. Furthermore, when users refer to the evaluation results to decide who to vote for, a comprehensive evaluation that includes emotional elements is provided, thereby enriching the information that supports the user's decision-making.

[0430] "Published minutes data" refers to record data of parliaments and meetings that are made public by governments and public institutions.

[0431] "Preprocessing" refers to processes such as data cleaning, tokenization, and removal of stop words that are carried out to prepare the acquired text data in a form that is easier to analyze.

[0432] "Natural language processing technology" is a technology that enables computers to understand and analyze human language, and is used for analyzing text and extracting information.

[0433] "Content of remarks" refers to the specific content and themes of remarks made by members of parliament in parliament or at meetings.

[0434] "Speech frequency" refers to the number of statements made by a member of parliament within a certain period of time.

[0435] "Attendance rate" is the percentage of times a member of parliament attended parliament or meetings.

[0436] "Quantitative and qualitative evaluation" means evaluation using both numerical data (quantitative evaluation) and qualitative aspects (qualitative evaluation).

[0437] A "legislator report card" is a report summarizing the results of an evaluation of a member of parliament's work, and is made public.

[0438] An "emotion engine" is a technology that analyzes emotions from text data and identifies emotions such as positive, negative, and neutral.

[0439] "Users" are general users who use the system to view the evaluation results of lawmakers.

[0440] This invention is a system that acquires publicly available minutes data and evaluates the performance of legislators using natural language processing technology and an emotion engine. A specific embodiment of this system is described below.

[0441] The server first obtains the publicly available meeting minutes data from the Internet. Specifically, it downloads the latest meeting minutes data using open data APIs provided by governments and public institutions. For example, it accesses the API using the Python requests library and obtains the data in JSON format.

[0442] The server then preprocesses the data, including cleaning the text, tokenizing it, and removing stop words. This is done using Python's pandas library and regular expressions (re library). For example, it removes unnecessary HTML tags and special characters, and splits the text into words using NLTK's word_tokenize function.

[0443] The preprocessed data is then analyzed using natural language processing technology. The server uses natural language processing libraries such as NLTK and spaCy to extract information such as the content of the legislator's remarks, their frequency, and their attendance rate. For example, topic modeling technology (e.g., LDA) is used to classify the content of the remarks, and legislators' remarks are categorized into topics such as "economy," "education," and "diplomacy."

[0444] The server then uses an emotion engine to analyze the emotions contained in the legislator's statements. Specifically, it uses the IBM Watson® Tone Analyzer API to classify statements as positive, negative, or neutral.

[0445] Based on these analysis results, the server evaluates the performance of the legislators quantitatively and qualitatively. The evaluation is carried out using indicators such as frequency of speech, attendance rate, and emotional balance. The evaluation results are generated as a "legislator report card" and made public through a web interface.

[0446] A user can check the evaluation of a specific legislator using a web browser. For example, if a user requests "Show me the report card of legislator A," the server retrieves the legislator A's evaluation data from the database and displays it on the web page.

[0447] Examples of prompts to input to a generative AI model include:

[0448] "Analyze the text data of the Diet minutes and extract the content of statements made by Diet members, their frequency of statements, and their attendance rates. Furthermore, use an emotion engine to analyze the emotions contained in the statements and evaluate the performance of Diet members quantitatively and qualitatively."

[0449] In this way, the system analyzes the minutes data and performs a series of processes to evaluate the performance of the members of parliament.

[0450] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0451] Step 1: Data Acquisition

[0452] The server retrieves publicly available meeting minutes data from the Internet. Specifically, it downloads the latest meeting minutes data using open data APIs provided by governments and public institutions. It accesses the API using the Python requests library and retrieves the data in JSON format. The input is the API endpoint URL, and the output is the retrieved meeting minutes data in JSON format.

[0453] Step 2: Data Preprocessing

[0454] The server preprocesses the acquired JSON-formatted meeting minutes data. Specifically, it converts the data into a data frame using Python's pandas library and removes unnecessary HTML tags and special characters using regular expressions (the re library). Next, it splits the text into words using NLTK's word_tokenize function and removes unnecessary words by referring to a stop word list. The input is the acquired JSON-formatted meeting minutes data, and the output is the preprocessed text data.

[0455] Step 3: Natural Language Processing Analysis

[0456] The server analyzes the preprocessed text data using natural language processing technology. Specifically, it uses natural language processing libraries such as NLTK and spaCy to extract information such as the content of the legislator's remarks, frequency of remarks, and attendance rate. For example, it uses topic modeling technology (e.g., LDA) to classify the content of remarks, categorizing legislator remarks into topics such as "economy," "education," and "diplomacy." The input is the preprocessed text data, and the output is the analysis results, such as the content of remarks, frequency of remarks, and attendance rate.

[0457] Step 4: Sentiment Analysis

[0458] The server uses an emotion engine to analyze the emotions contained in the legislator's statements. Specifically, it uses the IBM Watson Tone Analyzer API to classify statements as positive, negative, or neutral. The input is preprocessed text data, and the output is the result of the emotion analysis.

[0459] Step 5: Generate evaluation results

[0460] The server quantitatively and qualitatively evaluates the performance of legislators based on the results of natural language processing and sentiment analysis. Evaluations are made using indicators such as speech frequency, attendance rate, and emotional balance. The evaluation results are generated as a "legislator's report card." The inputs are the analysis results of speech content, speech frequency, attendance rate, and sentiment analysis results, and the output is the evaluation results in the form of a "legislator's report card."

[0461] Step 6: Publication of evaluation results

[0462] The server makes the generated "report card of the legislator" available to the public through a web interface. Users can check the evaluation of a specific legislator using a web browser. For example, if a user requests "Display the report card of legislator A," the server retrieves the evaluation data of legislator A from the database and displays it on the web page. The input is the user's request, and the output is the evaluation result displayed on the web page.

[0463] (Application example 1)

[0464] Next, a description will be given of Application Example 1 of Embodiment Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0465] Conventional lawmaker evaluation systems evaluated lawmakers based on quantitative data such as their speech content and attendance rate, but lacked the ability to provide personalized information that reflected the user's emotions and interests. They also lacked the ability to recommend lawmakers and topics that users might be interested in, making it difficult for users to efficiently obtain the information they needed. This limited their usefulness as reference information when deciding who to vote for.

[0466] 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.

[0467] In this invention, the server includes: means for AI to learn using published Diet minutes and the like as training data; means for the trained AI to quantitatively and qualitatively evaluate the performance of Diet members; means for creating and publishing the evaluation results as "Diet member report cards"; means including an emotion engine that recognizes the user's emotions; means for recommending Diet members and topics that the user may be interested in; and means for notifying the user of the latest statements and activities of Diet members they follow. This makes it possible to provide personalized information that reflects the user's emotions and interests, improving the usefulness of the information as reference when the user is deciding who to vote for.

[0468] "Published Diet minutes" are documents that record discussions and statements made in the Diet and are available to the public.

[0469] "Teacher data" is a dataset used for learning by a machine learning model, and is data that has been assigned a correct answer label.

[0470] "AI" is an abbreviation for artificial intelligence, a technology that allows computers to imitate human intelligence to learn and reason.

[0471] "Quantitative and qualitative evaluation" means evaluation based on numerical data (quantitative evaluation) and evaluation based on qualitative data (qualitative evaluation).

[0472] A "Member of Parliament Report Card" is a report summarizing the results of an evaluation based on the member's activities, statements, attendance rate, etc.

[0473] An "emotion engine" is a software or hardware system for recognizing and analyzing a user's emotions.

[0474] "Recommending politicians and topics that may be of interest to the user" means presenting relevant politicians and topics based on the user's interests and emotions.

[0475] "Notifying users of the latest statements and activities of politicians they follow" means providing users with real-time information on politicians they are interested in.

[0476] A system for carrying out this invention is configured as follows: The server includes means for AI to learn using published Diet minutes and the like as training data, means for the trained AI to quantitatively and qualitatively evaluate the performance of Diet members, means for creating and publishing the evaluation results as "Diet member report cards," means including an emotion engine that recognizes the user's emotions, means for recommending Diet members and topics that may interest the user, and means for notifying the user of the latest statements and activities of Diet members that the user follows.

[0477] Program processing explanation

[0478] The server first obtains the text data of the published Diet minutes and analyzes it using natural language processing technology. This analysis uses Hugging Face's Transformers library. As a result of the analysis, information such as the content of the Diet members' remarks, frequency of remarks, and attendance rate is extracted, and based on this, the performance of the Diet members is evaluated quantitatively and qualitatively. The evaluation results are compiled as a "Diet Member Report Card" and made publicly available for users to refer to.

[0479] Furthermore, the server uses an emotion engine to recognize the user's emotions. This emotion engine analyzes the user's input and behavioral data and recommends politicians and topics that the user may be interested in. This allows the user to efficiently obtain information based on their interests.

[0480] Users will be notified in real time of the latest statements and activities of the politicians they follow, allowing them to quickly keep up with the latest information on politicians they are interested in.

[0481] Specific examples

[0482] For example, if a user is interested in "environmental issues," the server will prioritize analyzing statements made by lawmakers on environmental issues and classify them into positive and negative. It will also recommend related lawmakers and topics based on the user's sentiment, allowing users to efficiently obtain information that matches their interests.

[0483] Prompt Sentence Examples

[0484] "Analyze the statements made by lawmakers on environmental issues and classify them into positive and negative. Also, recommend lawmakers that users might find interesting."

[0485] In this way, it becomes possible to provide personalized information that reflects the user's emotions and interests, improving the usefulness of the information as reference when the user decides who to vote for.

[0486] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0487] Step 1:

[0488] The server obtains the published text data of the Diet minutes. The input is the text data of the Diet minutes, and the output is text data in an analyzable format. Specifically, the server downloads the Diet minutes from a specified database or website and converts them into text format.

[0489] Step 2:

[0490] The server analyzes the acquired text data using natural language processing technology. The input is the text data acquired in step 1, and the output is information such as the content of each member's remarks, frequency of remarks, and attendance rate. Specifically, the server uses Hugging Face's Transformers library to analyze the text data and extract the content and frequency of each member's remarks.

[0491] Step 3:

[0492] The server evaluates the performance of the legislators quantitatively and qualitatively based on the analysis results. The input is the analysis results obtained in step 2, and the output is the legislator's evaluation score. Specifically, the server calculates a score based on the positive / negative evaluation of the content of the speech and attendance rate, and evaluates each legislator.

[0493] Step 4:

[0494] The server creates and publishes the evaluation results as "report cards for legislators." The input is the evaluation scores obtained in step 3, and the output is report card-style data that users can refer to. Specifically, the server organizes the evaluation scores and converts them into a format that can be viewed on a web page or application.

[0495] Step 5:

[0496] The server recognizes the user's emotions using an emotion engine. The input is the user's input data and behavioral data, and the output is the user's emotional state. Specifically, the server analyzes the user's text input and click history and uses the emotion engine to infer the user's emotional state.

[0497] Step 6:

[0498] The server recommends politicians and topics that the user is likely to be interested in. The input is the user's emotional state obtained in step 5 and the report card data obtained in step 4, and the output is the recommended politicians and topics. Specifically, the server selects relevant politicians and topics based on the user's emotional state and generates a recommendation list.

[0499] Step 7:

[0500] The server notifies the user of the latest statements and activities of the legislators they follow. The input is a list of legislators the user follows and the latest legislator activity data, and the output is a notification message. Specifically, the server references the follow list and notifies the user of relevant updates in real time.

[0501] Example 2

[0502] Next, a description will be given of Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0503] Conventional lawmaker evaluation systems typically base their evaluations on quantitative data such as the activities, statements, and attendance rate of lawmakers. However, these evaluations do not take into account the emotions and reactions of users, so the evaluation results do not necessarily reflect the opinions of voters. In addition, there are limited means for publishing the evaluation results, and they are often not provided in a format that is easily accessible to users. This creates the problem that the results are not fully utilized as reference information when voters decide who to vote for.

[0504] 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.

[0505] In this invention, the server includes means for an artificial intelligence to learn using published minutes and the like as training data, means for the trained artificial intelligence to quantitatively and qualitatively evaluate the performance of legislators, means for saving the evaluation results in a database, means for generating a report card in HTML format based on the saved evaluation results, means for publishing the generated report card on a website, means for recognizing the user's emotions, and means for reflecting the recognized emotions in the next evaluation results. This makes it possible to generate evaluation results that reflect the user's emotions and publish them in a form that users can easily access.

[0506] "Published minutes, etc." are documents and data that record the contents of assemblies and meetings and are made publicly available.

[0507] "Artificial intelligence" is a technology that mimics human intelligence through machine learning and data analysis.

[0508] "Quantitative and qualitative evaluation" means evaluating the value and performance of an object using numerical and qualitative data.

[0509] A "database" is a system for efficiently storing, managing, and retrieving data.

[0510] An "HTML report card" is a document in which the evaluation results are written using HTML (HyperText Markup Language) and formatted so that they can be displayed in a web browser.

[0511] A "website" is a collection of information that is publicly available on the Internet and accessible to users.

[0512] "Recognizing a user's emotions" means using sensors such as a camera and microphone to analyze the user's facial expressions and tone of voice and identify their emotions.

[0513] "Reflecting in the next evaluation result" means incorporating the recognized user emotion data into the next evaluation process to influence the evaluation result.

[0514] The present invention relates to a system for generating and publishing evaluation results of legislators. A specific embodiment of this system will be described below.

[0515] System configuration

[0516] Hardware

[0517] Server: manages the database, generates evaluation results, and hosts the website.

[0518] Terminal: The device used by the user to access the site (PC, smartphone, tablet, etc.).

[0519] Camera and microphone: sensors to recognize user emotions.

[0520] software

[0521] Database management system: MySQL, PostgreSQL, etc.

[0522] Web frameworks: Python, such as Django or Flask.

[0523] Emotion recognition library: OpenCV, TENSORFLOW (registered trademark), etc.

[0524] Program processing

[0525] Evaluation results stored in a database

[0526] The server receives the evaluation results of the legislator and stores them in a database such as MySQL or PostgreSQL. For example, if the evaluation results of legislator A are "Policy implementation ability: 8, Communication ability: 7, Reliability: 9," this is stored in the database.

[0527] HTML report card generation

[0528] The server retrieves the evaluation results from the database and generates an HTML report card using Django or Flask. The generated HTML file is published through a web server such as Apache (registered trademark) or Nginx. Users can view the report card through a browser.

[0529] User Emotion Recognition

[0530] When a user uses a device to view a legislator's report card, the device's camera and microphone are used to capture the user's facial expressions and tone of voice in real time. For example, if the user smiles, the device uses OpenCV to recognize the smile and TensorFlow to analyze the tone of voice.

[0531] Reflection of evaluation results by emotion engine

[0532] The device sends the analyzed emotion data to the server. The server stores this data in a database and reflects it when creating the next evaluation result. For example, if the data "User ID: 123, Congressman ID: A, Emotion: Positive" is sent, it will have a positive effect on the next evaluation result.

[0533] Specific examples

[0534] For example, if a user smiles while looking at Congressman A's report card, the device will recognize the smile using OpenCV and send it to the server as a positive emotion. The server will then reflect this emotion data in the next evaluation result, slightly increasing Congressman A's evaluation.

[0535] Prompt Sentence Examples

[0536] Examples of prompts to be input to a generative AI model include:

[0537] "Please analyze the user's emotions after seeing Congressman A's report card and reflect them in the next evaluation results. If the user smiles, please consider that emotion as positive and add it to the evaluation."

[0538] In this way, the system can reflect the user's feelings in the evaluation results and create a more accurate "report card for the legislator."

[0539] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0540] Step 1: Save the evaluation results in a database

[0541] The server receives the evaluation results of the legislator. As input, it receives the legislator's evaluation result data (e.g., "Policy implementation ability: 8, Communication ability: 7, Trustworthiness: 9"). The server stores this data in a database such as MySQL or PostgreSQL. Specifically, it executes an SQL query to insert the evaluation results into the "evaluation_results" table. As output, it obtains the evaluation results stored in the database.

[0542] Step 2: Generate an HTML report card

[0543] The server retrieves the evaluation results from the database. It executes a query (e.g., "SELECT FROM evaluation_results WHERE legislator ID = 'A'") with the evaluation results retrieved from the database as input. The server uses Python's Django or Flask to embed the evaluation results into an HTML template. Specifically, it uses a template engine to convert the evaluation results into HTML format. The generated HTML file is obtained as output. The server publishes this HTML file through a web server such as Apache or Nginx.

[0544] Step 3: Recognizing user emotions

[0545] A user uses a device to access a website and view a representative's report card. As input, the user's facial expressions and tone of voice are acquired in real time via a camera and microphone. The device then analyzes the user's emotions using libraries such as OpenCV and TensorFlow. Specifically, it uses a facial recognition algorithm to identify emotions such as smiling or anger, and a voice analysis algorithm to analyze the tone of voice. The analyzed emotional data (e.g., "Emotion: Positive") is obtained as output.

[0546] Step 4: Reflecting the evaluation results using the emotion engine

[0547] The device sends the analyzed emotion data to the server. As input, it sends emotion data (e.g., "User ID: 123, Congressman ID: A, Emotion: Positive"). The server stores this data in a database. Specifically, it executes an SQL query to insert the emotion data into the "user_emotions" table. The next time it creates an evaluation result, the server takes the stored emotion data into account and reflects it in the evaluation result. As output, it obtains a new evaluation result that reflects the emotion data.

[0548] In this way, the system can reflect the user's feelings in the evaluation results and create a more accurate "report card for the legislator."

[0549] (Application example 2)

[0550] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0551] Conventional lawmaker evaluation systems evaluate lawmakers based on quantitative and qualitative data such as their activities, statements, and attendance rates, but do not take into account the user's emotions and reactions. As a result, the evaluation results may not reflect the user's actual emotions and opinions, and users are not provided with sufficient useful information. Furthermore, in the shopping experience at physical stores, personalized product recommendations that take the user's emotions into account are not made, making it difficult to improve user satisfaction.

[0552] 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.

[0553] In this invention, the server includes a means for AI to learn using published Diet minutes and other data as training data, a means for the trained AI to quantitatively and qualitatively evaluate the performance of Diet members, a means for creating and publishing the evaluation results as "Diet member report cards," a means for recognizing users' emotions and reflecting the results in the evaluation results, and a means for recognizing users' emotions when viewing products and making personalized product recommendations based on that information. This makes it possible to provide Diet member evaluation results that reflect users' emotions and to make personalized product recommendations in physical stores.

[0554] "Published Diet minutes" are documents that record discussions and statements made in the Diet and are available to the public.

[0555] "Training data" is a set of input data and its correct labels used for learning a machine learning model.

[0556] "AI" is an abbreviation for artificial intelligence, a technology that allows computers to imitate human intelligence to learn and reason.

[0557] "Quantitative and qualitative evaluation" means evaluating an object using numerical and qualitative data.

[0558] A "legislator's report card" is a report summarizing the activities of a member of parliament and the results of their evaluation.

[0559] "Recognizing the user's emotions" means analyzing the user's emotions from their facial expressions, tone of voice, etc.

[0560] "Personalized product recommendation" refers to recommending products based on a user's individual preferences and emotions.

[0561] A "system" is a collection of devices and software that combine multiple elements to achieve a specific function.

[0562] A system for implementing the present invention is configured as follows.

[0563] System configuration

[0564] The system consists of a server, a user terminal (e.g., smart glasses), and software components such as an emotion recognition engine, a product recommendation engine, and a display module.

[0565] server

[0566] The server has a means for the AI ​​to learn from publicly available Diet minutes and other training data. The trained AI then quantitatively and qualitatively evaluates the performance of legislators, creating and publishing the results as "legislator report cards." It also has a means for recognizing users' emotions and reflecting those results in the evaluation results.

[0567] User terminal

[0568] The user terminal is a wearable device such as smart glasses, equipped with a camera and a display. It has a means of recognizing the user's emotions when viewing products and making personalized product recommendations based on that information.

[0569] Hardware and software used

[0570] Hardware: Smart glasses (camera, display), computer (built into the smart glasses)

[0571] Software: OpenCV (camera image acquisition and processing), EmotionRecognizer (emotion recognition engine), ProductRecommender (product recommendation engine), DisplayModule (display module)

[0572] Data processing and calculation

[0573] server

[0574] The server uses the published Diet minutes as training data for the AI ​​to learn from. After learning, the AI ​​evaluates the lawmakers based on data such as their activities, statements, and attendance rates. The evaluation results are generated as "lawmaker report cards" and published on a website. The server also recognizes the user's emotions and reflects the results in the next evaluation.

[0575] User terminal

[0576] The user device captures the user's facial expressions in real time through the smart glasses' camera and analyzes their emotions using EmotionRecognizer. The analyzed emotional data is sent to ProductRecommender, which then recommends products based on the user's emotions. Recommended product information is then displayed on the smart glasses' display.

[0577] Specific examples

[0578] If a user wears smart glasses in a physical store and expresses surprise when looking at a particular product, related new products and promotional information will be displayed based on that emotion.

[0579] Prompt Sentence Examples

[0580] "Develop an application for smart glasses that recognizes the emotions users feel when viewing products and provides personalized product recommendations based on those emotions. Include a feature to display new products and promotional information if the user is surprised."

[0581] The above is an embodiment of the present invention.

[0582] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0583] Step 1:

[0584] The server uses the published Diet minutes as training data to train the AI. The input is the text data of the Diet minutes, and the output is the trained AI model. Specifically, it analyzes the text data using natural language processing technology and extracts the statements and activities of Diet members as features.

[0585] Step 2:

[0586] The server uses the trained AI to quantitatively and qualitatively evaluate the performance of the legislators. The input is the legislator's activity data, and the output is the evaluation results. Specifically, the AI ​​model calculates a score based on data such as the legislator's remarks and attendance rate, and generates the evaluation results.

[0587] Step 3:

[0588] The server creates the evaluation results as "report cards for legislators" and publishes them on a website. The input is the evaluation results, and the output is the report card in HTML format. Specifically, the server saves the evaluation results in a database and generates an HTML file to display as a web page.

[0589] Step 4:

[0590] The user device captures the user's facial expressions in real time through the smart glasses' camera. The input is the camera image, and the output is the user's facial expression data. Specifically, the image captured by the camera is processed frame by frame, and facial expressions are analyzed using a face detection algorithm.

[0591] Step 5:

[0592] The user device analyzes the user's emotions using EmotionRecognizer. The input is the user's facial expression data, and the output is emotion data. Specifically, the facial expression data is input into an emotion recognition engine, which classifies emotions such as joy, anger, and surprise.

[0593] Step 6:

[0594] The user device sends the analyzed emotional data to the ProductRecommender, which then makes personalized product recommendations. The input is emotional data, and the output is a list of recommended products. Specifically, the system infers the user's preferences and interests based on the emotional data, and recommends related products.

[0595] Step 7:

[0596] The user terminal displays the recommended product information on the display of the smart glasses. The input is the recommended product list, and the output is the product information displayed on the display. Specifically, the recommended product list is input to the display module and displayed in a form that can be visually confirmed by the user.

[0597] The above are the specific processing steps for carrying out the present invention.

[0598] Example 3

[0599] Next, a description will be given of a third embodiment of the third embodiment. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0600] Conventional legislator evaluation systems mainly evaluate legislators based on quantitative data such as their activities and attendance rates, and are unable to reflect the emotions and opinions of users. As a result, when users use the evaluation results to decide who to vote for, their individual emotions and opinions are not reflected, reducing the reliability and usefulness of the evaluation results. Furthermore, the way the evaluation results are displayed can sometimes be difficult to understand visually, making it difficult for users to understand.

[0601] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[0602] In this invention, the server includes: means for an artificial intelligence to learn using publicly available parliamentary records and other training data; means for the trained artificial intelligence to quantitatively and qualitatively evaluate the performance of legislators; means for creating and publishing the evaluation results as a "legislator report card"; means for analyzing facial expressions, voice, and text input to recognize the user's emotions; means for reflecting the recognized emotional data in the evaluation results; and means for displaying the evaluation results in a form that can be referenced by the user. This makes it possible to provide evaluation results that reflect the user's emotions and opinions, improving the reliability and usefulness of users when using the evaluation results to decide who to vote for. Furthermore, displaying the evaluation results in a visually easy-to-understand format makes them easy for users to understand.

[0603] "Public parliamentary records" are documents containing data such as members' activities, statements, and attendance rates that can be obtained from official parliamentary websites and other public sources.

[0604] "Artificial intelligence" refers to computer systems that use techniques such as machine learning and deep learning to learn patterns from large amounts of data and automatically perform specific tasks.

[0605] "Quantitative and qualitative evaluation" means to comprehensively evaluate the work of legislators by combining evaluation based on numerical data (quantitative evaluation) with evaluation based on qualitative data (qualitative evaluation).

[0606] A "legislator report card" is a report summarizing the evaluation results of a member of parliament's activities, attendance rate, comments, etc., and is made publicly available for users to refer to.

[0607] "Recognizing user emotions" means identifying the emotions a user is feeling by analyzing their facial expressions, voice, text input, etc.

[0608] "Reflecting emotional data in the evaluation results" means adding or subtracting from the lawmaker's evaluation score based on the user's emotion recognition results, and reflecting this in the final evaluation results.

[0609] "Displaying the evaluation results in a form that users can refer to" means displaying the evaluation results on a website in a visually easy-to-understand format, such as a graph or table, so that users can easily understand them.

[0610] MODE FOR CARRYING OUT THE INVENTION

[0611] This invention is a system that provides voters with evaluation results for each legislator as reference information when deciding who to vote for. Specifically, voters refer to each legislator's report card on a website and decide who to vote for based on the evaluation results. It also uses an emotion engine to recognize the user's emotions and reflect them in the evaluation results.

[0612] Hardware and software used

[0613] 1. Server

[0614] Hardware: High-performance server machine

[0615] Software: Python, BeautifulSoup, pandas, OpenCV

[0616] 2. Terminal

[0617] Hardware: Your PC or smartphone

[0618] Software: Web browser, JavaScript (registered trademark), D3.js

[0619] 3. Users

[0620] Hardware: Webcam, microphone

[0621] Software: Web browser

[0622] Data processing and calculation

[0623] 1. Data Collection

[0624] The server uses Python's BeautifulSoup library to scrape parliamentary attendance data from official parliamentary websites, for example by visiting specific URLs and extracting attendance information from specific HTML tags.

[0625] 2. Data analysis

[0626] The server uses the Python pandas library to analyze the data and calculate an evaluation score for each legislator. For example, it implements an algorithm that gives a higher evaluation to legislators with an attendance rate of 90% or higher.

[0627] 3. Display of evaluation results

[0628] The device uses the JavaScript library D3.js to display the evaluation results in a graphical format, for example, displaying each legislator's evaluation score in a bar graph, allowing users to easily compare them.

[0629] 4. Emotion recognition

[0630] The emotion engine uses OpenCV to analyze the user's facial expressions and detect positive emotions. For example, if the user smiles through the webcam, it will recognize that smile as a positive emotion.

[0631] 5. Reflecting Emotional Data

[0632] If a user comments, "This legislator is doing a great job," the server adds that positive sentiment to the rating score. For example, the server implements an algorithm that gives additional points to legislators with many positive comments.

[0633] Specific examples

[0634] When a user accesses the website and views a member of parliament's report card, they can input a prompt sentence like the one below into the generative AI model to generate a sentence explaining the system's overview and processing procedures in natural language.

[0635] "Please explain the system that decides who to vote for based on the results of an evaluation of a member of parliament. Specifically, the system refers to each member of parliament's report card on a website and decides who to vote for based on the evaluation results. It also uses an emotion engine to recognize the user's emotions and reflect them in the evaluation results."

[0636] By inputting this prompt sentence into the generative AI model, a sentence explaining the system overview and processing procedure in natural language can be generated. The flow of the specific processing in Example 3 will be described with reference to FIG.

[0637] Program processing flow

[0638] Step 1: Data collection

[0639] Input: Official Congressional website URL

[0640] Output: Data on attendance rate and speech content of each member of parliament

[0641] Specific behavior:

[0642] The server uses Python's BeautifulSoup library to scrape the attendance data of members from the official parliamentary website. Specifically, it accesses specific URLs and extracts attendance information from specific HTML tags. For example, it parses a table listing each member's attendance rate to obtain the required data.

[0643] Step 2: Data analysis

[0644] Input: Data on attendance and speech content of each member of parliament collected in Step 1

[0645] Output: Evaluation score for each legislator

[0646] Specific behavior:

[0647] The server analyzes the data using Python's pandas library and calculates an evaluation score for each legislator. Specifically, it implements an algorithm that gives higher evaluations to legislators with attendance rates of 90% or higher. For example, it calculates an overall evaluation score based on attendance rate, number of comments, and quality of comments.

[0648] Step 3: View the evaluation results

[0649] Input: Evaluation scores for each legislator calculated in step 2

[0650] Output: Evaluation results displayed on the website

[0651] Specific behavior:

[0652] The device uses the JavaScript library D3.js to display the evaluation results in a graph format. Specifically, it displays each legislator's evaluation score in a bar graph, allowing users to easily compare them. For example, it generates a graph with the evaluation score on the vertical axis and the legislator's name on the horizontal axis.

[0653] Step 4: Emotion Recognition

[0654] Input: User facial expressions, voice, and text input

[0655] Output: User sentiment data

[0656] Specific behavior:

[0657] The emotion engine uses OpenCV to analyze the user's facial expressions and detect positive emotions. Specifically, if the user smiles through the webcam, it recognizes the smile as a positive emotion. For example, it analyzes facial feature points to detect smile patterns.

[0658] Step 5: Reflecting emotional data

[0659] Input: User emotion data obtained in step 4

[0660] Output: Updated rating score

[0661] Specific behavior:

[0662] When a user comments, "This legislator is doing a great job," the server adds that positive sentiment to the evaluation score. Specifically, the server implements an algorithm that gives additional points to legislators with many positive comments. For example, the server reflects the positive sentiment data in the evaluation score and updates the final evaluation result.

[0663] (Application example 3)

[0664] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0665] In conventional lawmaker evaluation systems, evaluation results are provided unilaterally and are unable to reflect the user's feelings or opinions. In addition, there are limited ways to check the evaluation results in real time, which means that users cannot fully utilize the information when deciding who to vote for.

[0666] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[0667] In this invention, the server includes: means for AI to learn using published Diet minutes and the like as training data; means for the trained AI to quantitatively and qualitatively evaluate the performance of Diet members; means for creating and publishing the evaluation results as "Diet member report cards"; means for analyzing the user's facial expressions, voice, text input, etc. using an emotion engine to recognize the user's emotions; means for reflecting the user's emotional information in the evaluation results; and means for the user to check the evaluation results in real time through an application installed on a smartphone. This makes it possible to provide evaluation results that reflect the user's emotions and opinions, and for the user to use them in real time as reference information when deciding who to vote for.

[0668] "Published Diet minutes" are documents that record the discussions and decisions made in the Diet and are available to the public.

[0669] "Training data" is a set of input data and its correct labels used for learning a machine learning model.

[0670] "AI" is an abbreviation for artificial intelligence, a technology that allows computers to imitate human intelligence to learn and reason.

[0671] "The work of a member of parliament" refers to the activities, statements, and attendance of members of parliament in the Diet and other official duties.

[0672] "Quantitative and qualitative evaluation" means that evaluation is conducted using both numerical and qualitative data.

[0673] A "Member of Parliament Report Card" is a report that evaluates a member of parliament's activities, comments, attendance, etc. and summarizes the results.

[0674] An "emotion engine" is software or algorithms that recognize and analyze user emotions.

[0675] "Facial expressions" refer to the movements and expressions of a user's face, and are one way of expressing emotions.

[0676] "Voice" refers to the voice or spoken words produced by the user, and is one of the means of expressing emotions.

[0677] "Text input" refers to textual information entered by a user using a keyboard or touchscreen.

[0678] A "smartphone" is a type of mobile phone and a multi-function device that can connect to the Internet and use applications.

[0679] "Real-time" refers to data and information being processed and provided to users immediately.

[0680] A system for implementing this invention is configured as follows: The server has a means for AI to learn from published Diet records and other data as training data. Specifically, the server collects Diet records and supplies them to a machine learning model as training data. This allows the AI ​​to learn about the activities and statements of Diet members and evaluate their performance quantitatively and qualitatively.

[0681] Next, the server has a means for the trained AI to quantitatively and qualitatively evaluate the performance of the legislators. The evaluation is calculated based on the legislators' activities, comments, attendance rate, etc. The evaluation results are compiled as "legislator report cards" and made public. This allows users to refer to the evaluation results and use them as reference information when deciding who to vote for.

[0682] Furthermore, the server has a means to analyze the user's facial expressions, voice, text input, etc. using an emotion engine to recognize the user's emotions. The emotion engine collects and analyzes the emotions expressed by the user regarding the results of the legislator's evaluation. For example, if a user comments, "This legislator is doing a great job," the text conveys the user's positive emotions. This information is also collected and analyzed by the emotion engine and reflected in the evaluation results.

[0683] The server also has a means to reflect the user's emotional information in the evaluation results, which allows the evaluation results to reflect the user's emotions and opinions, making it possible to provide more reliable information.

[0684] Finally, the server provides a means for users to check the evaluation results in real time through an application installed on their smartphones. Users can use their smartphones to check the evaluation results of lawmakers in real time and use this information as reference when deciding who to vote for.

[0685] As a concrete example, consider the case where a user opens the "Diet Member Evaluation Navigator" app on their smartphone to check the evaluation results of a specific legislator. When the user comments, "This legislator is doing a great job," and uploads a photo of their face, the app analyzes their emotions from the comment and photo and reflects this in the evaluation results. In this way, it is possible to provide evaluation results that reflect the user's emotions and opinions.

[0686] Examples of prompts for generative AI models include:

[0687] Open the "Assemblyman Evaluation Navigator" app and view the evaluation results for assemblyman ID "12345." Upload a user comment such as "This assemblyman is doing a great job," along with a photo of the person's face, and have the system analyze their emotions and reflect them in the evaluation results.

[0688] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[0689] Step 1:

[0690] The server collects published parliamentary records.

[0691] Input: Published Parliamentary Records

[0692] Output: Collected parliamentary records data

[0693] Specific operation: The server periodically crawls publicly available parliamentary minutes on the Internet and stores them in a database.

[0694] Step 2:

[0695] The server supplies the collected parliamentary minutes as training data to the machine learning model, allowing the AI ​​to learn.

[0696] Input: Collected parliamentary transcript data

[0697] Output: Trained AI model

[0698] Specific operation: The server preprocesses the collected parliamentary transcript data and feeds it to a machine learning algorithm to train a model.

[0699] Step 3:

[0700] The server uses a trained AI model to quantitatively and qualitatively evaluate the performance of lawmakers.

[0701] Input: Trained AI model, lawmaker activity data

[0702] Output: Evaluation results of legislators

[0703] Specific operation: The server inputs the activity data of the lawmakers into a trained AI model and generates evaluation results.

[0704] Step 4:

[0705] The server creates and publishes the evaluation results as "report cards for legislators."

[0706] Input: Evaluation results of the legislator

[0707] Output: "Congressman's report card"

[0708] Specific operation: The server formats the evaluation results and makes them available through a website or application.

[0709] Step 5:

[0710] The device uses an emotion engine to recognize the user's emotions by analyzing the user's facial expressions, voice, text input, etc.

[0711] Input: User facial expressions, voice, and text input

[0712] Output: User emotion data

[0713] Specific operation: The device uses a camera and microphone to capture the user's facial expressions and voice, receives text input, and analyzes it with an emotion engine.

[0714] Step 6:

[0715] The server reflects the user's emotional information in the evaluation results.

[0716] Input: User's emotional data, legislator evaluation results

[0717] Output: Evaluation results reflecting emotional information

[0718] Specific operation: The server integrates the user's emotional data into the evaluation results and generates evaluation results that reflect the emotional information.

[0719] Step 7:

[0720] The device allows users to check the evaluation results in real time through an application installed on their smartphone.

[0721] Input: Evaluation results that reflect emotional information

[0722] Output: Evaluation results for user review

[0723] Specific operation: The device displays the evaluation results through a smartphone app, allowing users to check them in real time.

[0724] 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.

[0725] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0726] Another example of generative AI is Gemini (registered trademark) (Internet search engine). <url: https: gemini.google.com ?hl="ja">) are listed.

[0727] 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.

[0728] [Second embodiment]

[0729] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0730] 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.

[0731] 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.

[0732] 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).

[0733] 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.

[0734] 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.

[0735] 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).

[0736] 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.

[0737] 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.

[0738] 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.

[0739] 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.

[0740] 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.

[0741] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[0742] "Example 1"

[0743] One embodiment of the present invention is a system in which AI learns from publicly available Diet minutes and other training data. Specifically, the system acquires text data from Diet minutes and analyzes it using natural language processing technology. The analysis extracts information such as the content and frequency of Diet members' remarks, attendance rate, etc., and uses this information to quantitatively and qualitatively evaluate the performance of Diet members.

[0744] "Example 2"

[0745] The evaluation results are compiled as "report cards for legislators" and made public. Specifically, the evaluation results for each legislator are stored in a database, and an HTML report card is generated based on the results. The generated report card is made public on a website so that voters can view it.

[0746] "Example 3"

[0747] Another embodiment of the present invention is a system that uses the evaluation results as reference information when voters decide who to vote for. Specifically, the system refers to each legislator's report card on a website and decides who to vote for based on the evaluation results. For example, it is possible to give priority to legislators with high attendance rates or high evaluations of the content of their statements.

[0748] The processing flow of each embodiment will be described below.

[0749] "Example 1"

[0750] Step 1: Obtain text data of the Diet minutes from the Internet.

[0751] Step 2: The acquired text data is analyzed using natural language processing technology. This analysis extracts information such as the content and frequency of the legislator's statements and attendance rate.

[0752] Step 3: Based on the extracted information, the AI ​​quantitatively and qualitatively evaluates the performance of the lawmakers.

[0753] "Example 2"

[0754] Step 1: Save the AI ​​evaluation results in a database.

[0755] Step 2: Generate an HTML report card based on the saved evaluation results.

[0756] Step 3: Publish the generated report card on a website so that voters can view it.

[0757] "Example 3"

[0758] Step 1: Voters view each representative's report card on the website.

[0759] Step 2: Voters decide who to vote for based on the evaluation results of the report card they have referenced. For example, they can prioritize lawmakers with high attendance rates or high evaluations of the content of their speeches.

[0760] Example 1

[0761] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0762] Conventional evaluation systems for legislators often require manual data collection and analysis to quantitatively and qualitatively evaluate legislators' performance, which is time-consuming and labor-intensive. Furthermore, evaluation results are not centrally managed and are not provided in an easily accessible format for users, making it difficult to use the evaluation results to decide who to vote for.

[0763] 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.

[0764] In this invention, the server includes: means for an artificial intelligence to learn using published minutes and the like as training data; means for the trained artificial intelligence to quantitatively and qualitatively evaluate the performance of legislators; means for creating and publishing the evaluation results as "legislator report cards"; means for acquiring text data of minutes via the Internet; means for preprocessing the acquired text data; means for analyzing the preprocessed data using natural language processing technology; means for evaluating the performance of legislators based on the analysis results; and means for outputting the evaluation results in a format that can be viewed by users. This makes it possible to efficiently and accurately evaluate the performance of legislators and provide the evaluation results in a format that users can easily refer to.

[0765] "Published minutes" are records of meetings that are made public by a government or public institution.

[0766] "Training data" is a dataset used for learning artificial intelligence.

[0767] "Artificial intelligence" is a technology that allows computer systems to behave in a way that mimics human intelligence.

[0768] "Quantitative evaluation" is a method of evaluating an object using numbers and statistics.

[0769] "Qualitative evaluation" is a method of evaluating an object by taking into account its qualitative aspects.

[0770] A "legislator report card" is a report summarizing the results of an evaluation of a member of parliament's work.

[0771] The "Internet" is a global communications network that interconnects computer networks around the world.

[0772] "Text data" is data that includes character information.

[0773] "Preprocessing" is the process of formatting and processing data before data analysis.

[0774] "Natural language processing technology" is a technology that uses computers to understand, analyze, and generate human language.

[0775] "Analysis" is the process of examining and evaluating data in detail.

[0776] "Evaluation results" are the results of evaluation obtained based on analysis.

[0777] "User" means a person or organization that uses this system.

[0778] "Format" is the specific way in which data or information is presented.

[0779] MODE FOR CARRYING OUT THE INVENTION

[0780] This invention is a system in which an artificial intelligence learns from publicly available minutes and other training data, and quantitatively and qualitatively evaluates the performance of legislators. A specific embodiment of this system will be described below.

[0781] 1. Program Generation

[0782] The server uses publicly available data such as meeting minutes as training data to build a generative AI model. This AI model is designed to analyze the text data in the minutes using natural language processing techniques. Specific software used includes Python's NLTK library and spaCy.

[0783] 2. Data Acquisition and Preprocessing

[0784] The terminal acquires the text data of the minutes via the Internet. The acquired data is sent to the server, which then performs the following preprocessing.

[0785] Text cleaning: Remove unnecessary symbols and spaces.

[0786] Tokenization: Splitting text into words and sentences.

[0787] Stop word removal: Remove meaningless words (e.g. "no", "wa", etc.).

[0788] 3. Data Analysis

[0789] The server analyzes the preprocessed data using natural language processing technology. Specifically, it performs the following analysis:

[0790] Speech analysis: Analyze the topic and sentiment of the speech.

[0791] Calculating speech frequency: Count the number of times each member of parliament speaks.

[0792] Calculating attendance rate: Extract the attendance rate of each member from the minutes.

[0793] 4. Generating ratings

[0794] The server evaluates the performance of the legislators based on the analysis results. The evaluation is performed as follows:

[0795] Quantitative evaluation: Quantify speaking frequency and attendance rate.

[0796] Qualitative evaluation: Evaluating the quality of what was said.

[0797] 5. Outputting the results

[0798] The server outputs the evaluation results in a format that can be viewed by the user, such as a web interface or a PDF report.

[0799] Specific examples

[0800] Specific use cases

[0801] If a user wants to evaluate the performance of a particular legislator, they can input a prompt like the following into the generative AI model:

[0802] Example prompt sentence:

[0803] "Based on the 2023 Diet minutes, please evaluate the content, frequency, and attendance rate of Congressman A."

[0804] When this prompt is entered, the following specific processing is performed:

[0805] 1. Data Acquisition:

[0806] The device will download the 2023 Diet minutes from the government's public API.

[0807] 2. Data preprocessing:

[0808] The server removes unnecessary symbols and spaces from the downloaded text data and performs tokenization.

[0809] The server removes stop words and converts the data into a format suitable for analysis.

[0810] 3. Data Analysis:

[0811] The server uses natural language processing technology to analyze the content of Congressman A's statements. For example, it detects that Congressman A frequently speaks about "economic policy."

[0812] The server calculates the frequency of Congressman A's statements and finds that the average number of times per month is 20.

[0813] The server calculates the attendance rate of Council Member A and extracts that it is 95%.

[0814] 4. Generate ratings:

[0815] The server quantitatively evaluates the performance of Councillor A based on frequency of speech and attendance rate.

[0816] The server qualitatively evaluates the performance of Congressman A based on the quality of his remarks.

[0817] 5. Result output:

[0818] The server generates the evaluation results as a PDF report that can be downloaded by the user.

[0819] In this way, the user can evaluate Congressman A's performance in detail.

[0820] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0821] Program processing flow

[0822] Step 1: Get the data

[0823] Input: The user enters a prompt to rate the performance of a particular legislator.

[0824] Operation: The device downloads text data of meeting minutes for a specified period from the government's public API via the Internet.

[0825] Output: Downloaded text data of the minutes.

[0826] Step 2: Preprocessing the data

[0827] Input: Text data of downloaded minutes.

[0828] Operation: The server performs the following preprocessing.

[0829] Text cleaning: Remove unnecessary symbols and spaces.

[0830] Tokenization: Splitting text into words and sentences.

[0831] Stop word removal: Remove meaningless words (e.g. "no", "wa", etc.).

[0832] Output: Preprocessed text data.

[0833] Step 3: Analyze the data

[0834] Input: Preprocessed text data.

[0835] Operation: The server performs the following analysis using natural language processing technology.

[0836] Speech analysis: Analyze the topic and sentiment of the speech.

[0837] Calculating speech frequency: Count the number of times each member of parliament speaks.

[0838] Calculating attendance rate: Extract the attendance rate of each member from the minutes.

[0839] Output: Analysis results (content of comments, frequency of comments, attendance rate).

[0840] Step 4: Generate ratings

[0841] Input: Analysis results (content of comments, frequency of comments, attendance rate).

[0842] How it works: The server evaluates the performance of the legislators based on the analysis results.

[0843] Quantitative evaluation: Quantify speaking frequency and attendance rate.

[0844] Qualitative evaluation: Evaluating the quality of what was said.

[0845] Output: Evaluation results (quantitative evaluation, qualitative evaluation).

[0846] Step 5: Output the results

[0847] Input: Evaluation results (quantitative evaluation, qualitative evaluation).

[0848] Behavior: The server outputs the evaluation results in a format that can be viewed by the user, such as a web interface or a PDF report.

[0849] Output: User-visible evaluation results.

[0850] Specific examples

[0851] Specific use cases

[0852] If a user wants to evaluate the performance of a particular legislator, they can input a prompt like the following into the generative AI model:

[0853] Example prompt sentence:

[0854] "Based on the 2023 Diet minutes, please evaluate the content, frequency, and attendance rate of Congressman A."

[0855] When this prompt is entered, the following specific processing is performed:

[0856] 1. Data Acquisition:

[0857] The device will download the 2023 Diet minutes from the government's public API.

[0858] 2. Data preprocessing:

[0859] The server removes unnecessary symbols and spaces from the downloaded text data and performs tokenization.

[0860] The server removes stop words and converts the data into a format suitable for analysis.

[0861] 3. Data Analysis:

[0862] The server uses natural language processing technology to analyze the content of Congressman A's statements. For example, it detects that Congressman A frequently speaks about "economic policy."

[0863] The server calculates the frequency of Congressman A's statements and finds that the average number of times per month is 20.

[0864] The server calculates the attendance rate of Council Member A and extracts that it is 95%.

[0865] 4. Generate ratings:

[0866] The server quantitatively evaluates the performance of Councillor A based on frequency of speech and attendance rate.

[0867] The server qualitatively evaluates the performance of Congressman A based on the quality of his remarks.

[0868] 5. Result output:

[0869] The server generates the evaluation results as a PDF report that can be downloaded by the user.

[0870] In this way, the user can evaluate Congressman A's performance in detail.

[0871] (Application example 1)

[0872] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0873] In traditional conferences and meetings, it was difficult to quantitatively and qualitatively evaluate participants' comments, frequency of comments, attendance rates, etc., resulting in problems that reduced the transparency and efficiency of the meetings. Furthermore, there was a lack of means to objectively evaluate participants' performance, making it impossible to use this information as reference for decision-making. This could result in a decline in the quality of meetings and a negative impact on the performance of the entire organization.

[0874] 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.

[0875] In this invention, the server includes a means for AI to learn using publicly available minutes and the like as training data, a means for the trained AI to quantitatively and qualitatively evaluate the performance of participants, a means for creating and publishing the evaluation results as a "participant report card," and a means for improving the transparency and efficiency of meetings. This makes it possible to objectively evaluate the content and frequency of participants' comments, attendance rates, etc., and improve the transparency and efficiency of meetings.

[0876] A "published minutes" is a document that records the contents of a conference or meeting and is available to the public.

[0877] "Training data" is a dataset used to train a machine learning model.

[0878] "AI" is an abbreviation for artificial intelligence, a system that mimics human intellectual tasks using technologies such as machine learning and natural language processing.

[0879] A "participant" is someone who attends a conference or meeting and expresses their opinions or comments.

[0880] "Quantitative evaluation" is a method of evaluation using numerical and statistical data.

[0881] "Qualitative evaluation" is a method of evaluation that takes into account qualitative aspects.

[0882] A "participant report card" is a document of evaluation results created based on the content and frequency of participants' comments, attendance rate, etc.

[0883] "Transparency" means that information is not hidden but is made public and available for anyone to see.

[0884] "Efficiency" is the state in which the time and effort required to achieve a goal is minimized.

[0885] A system for carrying out this invention is configured as follows: The server includes means for AI to learn using publicly available minutes and the like as training data, means for the trained AI to quantitatively and qualitatively evaluate the performance of participants, means for creating and publishing the evaluation results as "participant report cards," and means for improving the transparency and efficiency of meetings.

[0886] The server analyzes the minutes text using Python and the natural language processing library spaCy. Specifically, the server performs the following steps:

[0887] First, the server reads the minutes text from a file. Next, it analyzes the read text using spaCy to extract the content of each speaker's comments. The extracted content is categorized by speaker and the frequency of each comment is calculated. This allows for a quantitative evaluation of the content and frequency of each participant's comments.

[0888] The evaluation results are compiled as a "participant report card" and made publicly available for users to refer to, allowing them to use the information to improve the transparency and efficiency of their meetings.

[0889] As a concrete example, let's say we have the following meeting minutes:

[0890] Tanaka: Let's discuss a new project today.

[0891] Suzuki: Before that, let's check the minutes of the last meeting.

[0892] Tanaka: Understood. I will check the minutes.

[0893] An analysis of the minutes reveals the following results:

[0894] Tanaka: 2 comments

[0895] Suzuki: 1 statement

[0896] An example of a prompt sentence to input to the generative AI model is as follows:

[0897] Analyze the following meeting minutes and calculate the frequency of each speaker's comments.

[0898] Tanaka: Let's discuss a new project today.

[0899] Suzuki: Before that, let's check the minutes of the last meeting.

[0900] Tanaka: Understood. I will check the minutes.

[0901] In this way, the server can analyze the content and frequency of participants' comments and publish the evaluation results as a "participant report card," thereby improving the transparency and efficiency of meetings.

[0902] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0903] Step 1:

[0904] The server reads the published minutes text from a file. The input is the minutes text file, and the output is text data. This text data serves as the basis for subsequent natural language processing.

[0905] Step 2:

[0906] The server analyzes the loaded text data using spaCy, a natural language processing library. The input is the text data obtained in step 1, and the output is a parsed document object. This document object contains information such as sentence and word tokenization, part-of-speech tagging, and dependency analysis.

[0907] Step 3:

[0908] The server extracts the content of each speaker's comments from the analyzed document object. The input is the document object obtained in step 2, and the output is a list of comments categorized by speaker. Specifically, the speaker and the content of the comments are separated based on the ":" in the sentence, and the list is organized by speaker.

[0909] Step 4:

[0910] The server calculates the frequency of each speaker's comments. The input is the list of comments obtained in step 3, and the output is a dictionary of the frequency of each speaker's comments. Specifically, the server counts the number of comments made by each speaker and saves them in dictionary format.

[0911] Step 5:

[0912] The server creates a "participant report card" based on the dictionary of comment frequency. The input is the dictionary of comment frequency obtained in step 4, and the output is the evaluation results as a "participant report card." Specifically, the evaluation of each participant is quantitatively described based on the comment frequency and saved in document format.

[0913] Step 6:

[0914] The server publishes the created "Participant Report Cards" in a form that users can refer to. The input is the evaluation results obtained in step 5, and the output is a web page or dashboard that users can access. Specifically, the evaluation results are converted into HTML format and published on the web server.

[0915] In this way, the server analyzes the minutes text, evaluates the content and frequency of participants' comments, and publishes the results as a "participant report card," thereby improving the transparency and efficiency of meetings.

[0916] Example 2

[0917] Next, a description will be given of Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0918] In the conventional system for evaluating legislators, the evaluation results are not managed centrally, and the evaluations lack transparency and reliability. In addition, the evaluation results are not properly disclosed to the general public, which creates the problem that voters cannot use them as reference information when deciding who to vote for.

[0919] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes a means for an artificial intelligence to learn using published minutes and the like as training data, a means for the trained artificial intelligence to quantitatively and qualitatively evaluate the performance of the legislators, a means for saving the evaluation results in a database, a means for generating a report card in HTML format based on the saved evaluation results, and a means for publishing the generated report card on a website. This enables centralized management of evaluation results and ensures transparency, allowing voters to decide who to vote for based on reliable information.

[0920] "Published minutes, etc." are documents and data that record the content of statements and discussions made in official forums such as parliaments and conferences and are made publicly available.

[0921] "Artificial intelligence" is a technology that enables computer systems to mimic human intelligence and perform intellectual tasks such as learning, reasoning, and recognition.

[0922] The "means of learning" refers to the process by which AI analyzes data, such as publicly available minutes, to find patterns and rules.

[0923] "Quantitative and qualitative evaluation methods" refers to the process by which artificial intelligence evaluates the performance of legislators using numerical data (quantitative) and qualitative data (qualitative).

[0924] "Evaluation results" are data obtained as a result of AI evaluating the performance of lawmakers, and include the activities, statements, attendance rate, etc. of lawmakers.

[0925] A "database" is a digital information management system for efficiently storing, managing, and searching evaluation results.

[0926] An "HTML report card" is a web page document created based on the evaluation results to visually display the evaluation of a member of parliament.

[0927] "Publication on a website" is the process by which the generated HTML report card is made accessible to the public on the Internet.

[0928] "Users" are ordinary voters and users who access the website and view the report cards of their representatives.

[0929] This invention relates to a system for evaluating the performance of assembly members and publishing the evaluation results as a “asset member report card.” A specific embodiment of this system will be described below.

[0930] 1. Data collection and storage

[0931] The server collects publicly available minutes and other training data, including official records of parliaments and meetings. The collected data is stored in a MySQL database. Specifically, the evaluation results are inserted into the database using the INSERT statement.

[0932] 2. Data processing

[0933] The server retrieves the evaluation results from the MySQL database. Specifically, it extracts the required data using a SELECT statement. The retrieved data is converted into a data frame using the Python Pandas library and formatted as needed.

[0934] 3. Report Card Generation

[0935] The server uses the Jinja2 template engine to generate an HTML report card based on the processed data. Specifically, it generates the HTML by embedding the data in a template file. The generated HTML file is saved in the public directory of the web server.

[0936] 4. Publication of report cards

[0937] The server uses the Flask framework to build a web server. An endpoint for displaying report cards is set up and made accessible to users. A user accesses the website and views the report card of a specific legislator. Specifically, by entering the legislator's name in the search bar and clicking the search button, the report card of the corresponding legislator is displayed.

[0938] Specific examples

[0939] For example, consider a case where a user accesses a website and wants to view the report card of a particular politician. The user enters the politician's name in the website's search bar and clicks the search button. The server retrieves the evaluation results of the relevant politician from the database and generates and displays the report card in HTML format.

[0940] Prompt Sentence Examples

[0941] "Write a program that generates an HTML report card based on the evaluation results of a specific council member and publishes it on a website. The evaluation results are stored in a MySQL database, and you will use Python's Pandas library and the Jinja2 template engine to process the data and generate the report card. You will then publish the report card on the website using the Flask framework."

[0942] The above is a specific embodiment for carrying out the present invention.

[0943] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0944] Step 1:

[0945] The server collects publicly available minutes and other data as training data. Specifically, it obtains official records of parliaments and meetings through APIs and manual input forms. Input data includes what council members said, attendance rates, and activities. This data is stored in a MySQL database. Specifically, it inserts the evaluation results into the database using INSERT statements. The output is the evaluation results stored in the database.

[0946] Step 2:

[0947] The server retrieves the evaluation results from the MySQL database. Specifically, it uses a SELECT statement to extract the necessary data. The input is the evaluation results stored in the database. The retrieved data is converted into a data frame using the Python Pandas library. The converted data frame is then formatted as needed. For example, the evaluation scores are sorted in ascending order. The output is a formatted data frame.

[0948] Step 3:

[0949] The server uses the Jinja2 template engine to generate an HTML report card based on the processed data. The input is a formatted data frame. Specifically, HTML is generated by embedding data in a template file. The generated HTML file is saved in the public directory of the web server. The output is the generated HTML report card.

[0950] Step 4:

[0951] The server uses the Flask framework to build a web server. An endpoint for displaying report cards is set up. The input is the generated report card in HTML format. A user accesses the website and wants to view the report card of a specific legislator. Specifically, they enter the legislator's name in the search bar and click the search button. The server receives the user's request and displays the report card of the relevant legislator. The output is the report card displayed in the user's browser.

[0952] (Application example 2)

[0953] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0954] In conventional evaluation systems, it was difficult for users to easily refer to evaluation results based on the target's activities, comments, attendance rate, etc., and use them as a reference for decision-making. In addition, there was a lack of a way to save the evaluation results in a database and publish them in HTML format, making it difficult for users to visually check the evaluation results and post reviews.

[0955] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes a means for the artificial intelligence to learn using published minutes and the like as training data, a means for the learned artificial intelligence to quantitatively and qualitatively evaluate the target's performance, a means for creating and publishing the evaluation results as a "report card," a means for saving the evaluation results in a database and generating the report card in HTML format, and a means for users to refer to the evaluation results and post reviews. This allows users to visually check the evaluation results and use them as a reference for decision-making.

[0956] "Published minutes, etc." refers to minutes and related documents that are made publicly accessible.

[0957] "Artificial intelligence" refers to systems that mimic human intelligence through machine learning and data analysis.

[0958] "Means of learning" refers to the process by which artificial intelligence analyzes data based on publicly available minutes and other documents to find patterns and rules.

[0959] "Quantitative and qualitative evaluation methods" refers to the process of evaluating the functioning of an object using numerical and qualitative data.

[0960] A "report card" refers to a report-style document summarizing evaluation results.

[0961] "Disclosure means" refers to the process of making evaluation results publicly accessible.

[0962] "Means for storing in a database" refers to the process of storing the evaluation results in a structured format in a database.

[0963] "Means for generating an HTML report card" refers to the process of generating the evaluation results as an HTML document.

[0964] "Means for users to view rating results and post reviews" refers to an interface that allows users to view rating results and add their own opinions and ratings.

[0965] To implement the present invention, the following system configuration is required.

[0966] System Configuration

[0967] 1. Server

[0968] Hardware: A server with a powerful processor, sufficient memory, and storage.

[0969] Software: Python, Flask, SQLite, generative AI models.

[0970] 2. Terminal

[0971] Hardware: Smartphone.

[0972] Software: Web browser or dedicated application.

[0973] 3. Users

[0974] Role: View ratings and post reviews.

[0975] Program processing explanation

[0976] server

[0977] The server provides a means for the AI ​​to learn from publicly available meeting minutes and other data as training data. Specifically, the generative AI model is used to analyze the meeting minutes data and quantitatively and qualitatively evaluate the target's performance. The evaluation results are stored in a database and generated as a report card in HTML format. This allows users to visually check the evaluation results and use them as a reference for decision-making.

[0978] Terminal

[0979] The terminal provides an interface that allows users to view the evaluation results and post reviews. Using their smartphones, users can view the HTML report card provided by the server and add their own opinions and evaluations.

[0980] Specific examples

[0981] For example, when a user opens the app on their smartphone, they are presented with a list of nearby stores' ratings. Each store's rating includes reviews and star ratings posted by other users. Users can post new reviews, which are then sent to the server and stored in a database.

[0982] Prompt Sentence Examples

[0983] "Create a smartphone app that displays the ratings of nearby businesses and allows users to post reviews. The ratings will be stored in a database and displayed in HTML format."

[0984] The above is a specific embodiment for carrying out the present invention.

[0985] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0986] Step 1:

[0987] The server collects published minutes and stores them in a database. The input is the published minutes and other data, and the output is the raw data stored in the database. Specifically, the minutes data is obtained using web scraping and APIs and stored in an SQLite database.

[0988] Step 2:

[0989] The server uses a generative AI model to analyze the minutes data stored in the database and quantitatively and qualitatively evaluate the subject's performance. The input is the minutes data stored in the database, and the output is the evaluation results. Specifically, it uses natural language processing technology to analyze the minutes data and evaluate each subject's activities, comments, attendance rate, etc.

[0990] Step 3:

[0991] The server saves the evaluation results in a database and generates a report card in HTML format. The input is the evaluation results, and the output is an HTML report card. Specifically, it generates an HTML template based on the evaluation results and saves it in the database.

[0992] Step 4:

[0993] The server publishes the report card in HTML format through a web server so that users can view the evaluation results. The input is the report card in HTML format, and the output is the report card displayed in the user's browser. Specifically, a web server is built using Flask so that users can access it.

[0994] Step 5:

[0995] Users access the server using a smartphone browser or a dedicated app to view the evaluation results. The input is the user's access request, and the output is a display of the evaluation results. Specifically, the user opens a browser and accesses the server's URL, and the report card is displayed.

[0996] Step 6:

[0997] Users post reviews based on the evaluation results. The input is the user's review content, and the output is the review sent to the server. Specifically, the user enters the review in the form and presses the submit button, which sends the review to the server.

[0998] Step 7:

[0999] The server stores the reviews sent by users in a database and makes them available for other users to refer to. The input is the user's review content, and the output is the review saved in the database. Specifically, the review content is stored in the database and reflected in an HTML report card.

[1000] The above are the specific processing steps for carrying out the present invention.

[1001] Example 3

[1002] Next, a description will be given of Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1003] In conventional elections, it is difficult for voters to fully understand the activities and evaluations of their representatives, resulting in a lack of information to decide who to vote for. In addition, there is a lack of means to effectively utilize publicly available parliamentary records and quantitatively and qualitatively evaluate the activities of representatives, making it difficult for voters to decide who to vote for based on reliable information.

[1004] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[1005] In this invention, the server includes means for the artificial intelligence to learn using publicly available parliamentary records and the like as training data, means for the trained artificial intelligence to quantitatively and qualitatively evaluate the activities of legislators, means for creating and publishing the evaluation results as "legislator report cards," means for displaying the evaluation results on a website, and means for users to refer to the evaluation results on the website and decide who to vote for. This enables voters to evaluate the activities of legislators based on reliable information and decide who to vote for.

[1006] "Public parliamentary records" are information such as minutes, attendance records, and statements made that are officially recorded in parliament and made available to the public.

[1007] "Artificial intelligence" is a system that uses machine learning and data analysis techniques to learn patterns from data and make predictions and classifications.

[1008] "The activities of a member of parliament" are the actions and duties of a member of parliament, such as attending parliament, speaking, proposing bills, and voting.

[1009] "Quantitative and qualitative evaluation" means conducting both quantitative evaluation using numerical data and qualitative evaluation that evaluates content and quality.

[1010] A "legislator report card" is a report summarizing the activities and evaluation results of a member of parliament, and is a resource that helps voters understand the performance of their members.

[1011] A "website" is an online page published on the Internet that users can access and view information.

[1012] A "user" is a voter who uses the system to obtain information and decide how to vote.

[1013] "Training data" is a dataset consisting of known input and output pairs that is used to train an artificial intelligence.

[1014] "Learning" is the process by which artificial intelligence finds patterns and rules based on training data, improving the accuracy of predictions and classifications.

[1015] The "evaluation results" are the results of the AI ​​analyzing the activities of lawmakers and evaluating them quantitatively and qualitatively.

[1016] This invention is a system that provides evaluation results that voters can use as a reference when deciding who to vote for. Specifically, a server collects publicly available parliamentary records, evaluates the activities of legislators using artificial intelligence, and displays the evaluation results on a website. Users can refer to the evaluation results on the website and decide how to vote appropriately.

[1017] Hardware and software used

[1018] 1. Server

[1019] The server collects data, processes the data, generates evaluation results, and displays them on the website.

[1020] Software used includes Python, BeautifulSoup, Pandas, Scikit-learn, Flask or Django.

[1021] 2. Terminal

[1022] The terminal is used by the user to access the website and view the evaluation results.

[1023] The hardware used includes computers, smartphones, tablets, etc.

[1024] The software used includes a web browser (such as Google® Chrome, Mozilla Firefox, or Safari).

[1025] 3. Users

[1026] Users access the website as voters and decide who to vote for based on the evaluation results.

[1027] Program processing

[1028] 1. Data Collection

[1029] The server collects data such as each member's attendance rate and speech content from the parliament's official website and publicly available minutes.

[1030] Example: The server uses Python's BeautifulSoup library to scrape HTML data from the official parliamentary website, for example, parsing the page listing the attendance rates of members of parliament and extracting the required data.

[1031] 2. Data processing

[1032] The server organizes the collected data and creates a report card for each member of parliament, including attendance rate, evaluation of speech content, number of bills proposed, and more.

[1033] Example: The server uses the Pandas library to convert the collected data into a data frame. Then, it extracts the necessary information for each member of parliament and formats it into a report card format. For example, it extracts the attendance column and stores it in a new data frame along with each member's name.

[1034] 3. Generating evaluation results

[1035] The server generates evaluation results for each member of parliament, which are scored based on their attendance rate and the content of their comments.

[1036] Example: The server uses the Scikit-learn library to build a model for scoring the assessment results, for example a linear regression model that assigns higher scores to students with higher attendance rates.

[1037] 4. Display on the Website

[1038] The server displays the generated evaluation results on a website, and users can access the website to view the report cards of each member of parliament.

[1039] Example: The server uses a web framework such as Flask or Django to create a web page that displays the evaluation results. For example, it generates an HTML page that displays the names of each legislator and their evaluation scores in a table.

[1040] 5. User voting decision

[1041] Users can refer to the report cards of each member of parliament on the website and decide who to vote for based on the evaluation results.

[1042] Example: A user accesses a website using a browser, checks the displayed ratings, and, for example, selects the legislator with the highest rating score and decides to vote for that legislator.

[1043] Prompt Sentence Examples

[1044] Examples of prompts to be input to a generative AI model include:

[1045] "In order to decide which representative to vote for in the next election, I would like to refer to a report card based on the attendance rate and comments of each representative. I would like to check the evaluation results on the website and choose the representative with the highest evaluation."

[1046] By using this prompt sentence, the generative AI model can understand what information the user is looking for and provide an appropriate evaluation result. The flow of the identification process in Example 3 will be described with reference to Figure 15.

[1047] Step 1:

[1048] Data collection

[1049] The server collects data such as each member's attendance rate and speech content from the parliament's official website and publicly available minutes.

[1050] Input: URL of the official parliament website or publicly available minutes.

[1051] What it does: The server uses Python's BeautifulSoup library to scrape HTML data from the official parliamentary website. For example, it parses the page listing the attendance rates of members of parliament and extracts the necessary data.

[1052] Output: Raw data such as attendance rate and speech content of each member of parliament.

[1053] Step 2:

[1054] Data Processing

[1055] The server organizes the collected data and creates a report card for each member of parliament, including attendance rate, evaluation of speech content, number of bills proposed, and more.

[1056] Input: The raw data collected in step 1.

[1057] What it does: The server uses the Pandas library to convert the collected data into a data frame. Then, it extracts the necessary information for each member of parliament and formats it into a report card format. For example, it extracts the attendance column and stores it in a new data frame along with each member's name.

[1058] Output: Report card data for each member of parliament.

[1059] Step 3:

[1060] Generating evaluation results

[1061] The server generates evaluation results for each member of parliament, which are scored based on their attendance rate and the content of their comments.

[1062] Input: Report card data created in step 2.

[1063] What it does: The server uses the Scikit-learn library to build a model for scoring the assessment results, for example, a linear regression model that assigns higher scores to students with higher attendance rates.

[1064] Output: Evaluation scores for each legislator.

[1065] Step 4:

[1066] Display on the website

[1067] The server displays the generated evaluation results on a website, and users can access the website to view the report cards of each member of parliament.

[1068] Input: The rating scores generated in step 3.

[1069] What it does: The server uses a web framework such as Flask or Django to create a web page that displays the evaluation results. For example, it generates an HTML page that displays the names of each legislator and their evaluation scores in a table.

[1070] Output: A web page that displays the evaluation results.

[1071] Step 5:

[1072] User voting decision

[1073] Users can refer to the report cards of each member of parliament on the website and decide who to vote for based on the evaluation results.

[1074] Input: The assessment results displayed on the website.

[1075] Specific operation: A user accesses a website using a browser, checks the displayed evaluation results, and, for example, selects the legislator with the highest evaluation score and decides to vote for that legislator.

[1076] Output: User's vote decision.

[1077] (Application example 3)

[1078] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1079] In conventional lawmaker evaluation systems, evaluation results are simply published on a website, and users lack sufficient interactivity and convenience when deciding who to vote for based on that information. Furthermore, there is a lack of means to provide information in physical stores, and there is no system in place to allow users to check evaluation results in real time and receive recommendations on the optimal candidate to vote for.

[1080] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[1081] In this invention, the server includes means for AI to learn using published Diet minutes and the like as training data, means for the trained AI to quantitatively and qualitatively evaluate the performance of Diet members, means for creating and publishing the evaluation results as "Diet member report cards," means for linking with interactive displays in physical stores and displaying information on smartphones using QR codes, means for recommending optimal voting destinations to users based on the evaluation results, and means for updating the evaluation results and recommendation information in real time. This enables users to interactively check the evaluation results in physical stores and receive optimal voting destination recommendations in real time.

[1082] "Published Diet minutes" are documents that record the discussions and statements made in the Diet and are available to the public.

[1083] "Teacher data" is a dataset used for learning by a machine learning model, and is data that has been assigned a correct answer label.

[1084] "AI" is an abbreviation for artificial intelligence, a technology that allows computers to imitate human intelligence to learn and reason.

[1085] "The work of a member of parliament" is a general term for the activities, statements, attendance, etc. that members of parliament undertake in the Diet and other official duties.

[1086] "Quantitative and qualitative evaluation" means evaluation using both numerical data (quantitative) and qualitative data (qualitative).

[1087] A "Member of Parliament Report Card" is a report summarizing the evaluation results of a member of parliament's activities, comments, attendance rate, etc.

[1088] An "interactive display in a physical store" is a display device that can be directly operated by a user and is installed in a physical store.

[1089] A "QR code" is a type of two-dimensional barcode that can be read by devices such as smartphones.

[1090] A "smartphone" is a mobile phone that can connect to the Internet and use applications.

[1091] "Recommending the best voting option" means suggesting the most suitable voting option for the user.

[1092] "Real-time updates" means that data and information are updated immediately and always kept up to date.

[1093] The system for implementing this invention is configured as follows: First, the server has a means for the AI ​​to learn from publicly available Diet minutes and other training data. This AI is used to quantitatively and qualitatively evaluate the performance of Diet members. The evaluation results are compiled as "Diet member report cards" and made public.

[1094] The second step involves linking the system to interactive displays in physical stores and displaying information on smartphones using QR codes. Users can check the evaluation results of each legislator by scanning the QR code displayed on the display in the store with their smartphone.

[1095] Furthermore, the server has a means to recommend the best voting destination to the user based on the evaluation results. This recommendation is made in real time based on the evaluation result data. The evaluation results and recommendation information are always kept up to date.

[1096] The specific hardware used is a smartphone and an interactive display in a physical store. The software used is Flask (a Python web framework) and requests (a library for sending HTTP requests). Data processing involves obtaining the evaluation results of each legislator in JSON format and sorting them to recommend the optimal voting option to the user. Data calculation involves executing recommendation logic based on the evaluation results of attendance rate, speech content, and policy performance.

[1097] For example, when a user scans a QR code displayed on a display in a physical store, the evaluation results for each politician are displayed on their smartphone. Based on this information, the user can confirm who the application recommends for voting.

[1098] An example of a prompt is, "Please recommend the best candidate to vote for based on your evaluation of each member of parliament's attendance rate, speech content, and policy achievements."

[1099] In this way, users can interactively check the evaluation results in a physical store and receive recommendations for the optimal voting destination in real time.

[1100] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[1101] Step 1:

[1102] The server collects publicly available Diet minutes and other data as training data. It receives Diet minutes data as input and prepares it as training data for the AI ​​model. It processes the data by cleaning the text data and adding necessary labels. It outputs a dataset in a format that the AI ​​model can learn from.

[1103] Step 2:

[1104] The server trains the AI ​​model using the collected training data. The dataset generated in step 1 is used as input. As data calculation, the AI ​​model learns patterns to quantitatively and qualitatively evaluate the performance of legislators. The output is a trained AI model.

[1105] Step 3:

[1106] The server evaluates the performance of lawmakers using a trained AI model. As input, it receives the latest parliamentary minutes data and inputs this into the AI ​​model. The data is then calculated to evaluate the lawmakers' attendance rates, statements, policy achievements, etc. The output is the evaluation results for each lawmaker.

[1107] Step 4:

[1108] The server creates the evaluation results as a "report card for the legislators" and publishes them on a website. The evaluation results obtained in step 3 are used as input. The data is processed by formatting the evaluation results into an easy-to-read format. The output is a web page that can be viewed by users.

[1109] Step 5:

[1110] The user scans a QR code displayed on an interactive display in a physical store with their smartphone. As input, the smartphone reads the information in the QR code. Specifically, the smartphone's camera is used to scan the QR code. As output, the smartphone obtains the URL of the page displaying the evaluation results.

[1111] Step 6:

[1112] The terminal (smartphone) accesses the server using the URL obtained from the QR code and displays the evaluation results. The URL obtained from the QR code is used as input. As data calculation, the evaluation result data is obtained from the server and displayed on the smartphone screen. As output, a screen is provided where the user can check the evaluation results.

[1113] Step 7:

[1114] The server recommends the optimal voting destination to the user based on the evaluation results. The evaluation results viewed by the user are used as input. The data calculation involves sorting the evaluation results and calculating the optimal voting destination. The output shows the voting destination recommended to the user.

[1115] Step 8:

[1116] The server updates the evaluation results and recommendation information in real time. It receives new Diet minutes data and user feedback as input. It processes the data by retraining the AI ​​model with the latest data. It always provides the latest evaluation results and recommendation information as output.

[1117] 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.

[1118] "Example 1"

[1119] One embodiment of the present invention is a system that incorporates an emotion engine. In this system, AI learns from publicly available Diet records and other training data, and the trained AI quantitatively and qualitatively evaluates the performance of Diet members. The evaluation results are compiled and made public as a "Diet member report card." Furthermore, this system includes an emotion engine that recognizes the user's emotions.

[1120] "Example 2"

[1121] The emotion engine recognizes the user's emotions and reflects them in the evaluation results. Specifically, the system captures the user's reaction when they view the lawmaker's report card via camera and microphone, and analyzes their emotions from their facial expressions and tone of voice. For example, if a user feels happy, angry, or surprised when they see the lawmaker's evaluation results, that emotion is recognized by the emotion engine and reflected in the next evaluation results.

[1122] "Example 3"

[1123] The emotion engine also analyzes the user's facial expressions, voice, and text input to recognize their emotions. For example, if a user comments on social media in response to the evaluation results of a lawmaker, "This lawmaker is doing a great job," the text conveys the user's positive emotions. This information is also collected and analyzed by the emotion engine and reflected in the evaluation results.

[1124] The processing flow of each embodiment will be described below.

[1125] "Example 1"

[1126] Step 1: The AI ​​learns from publicly available parliamentary records and other training data.

[1127] Step 2: The trained AI evaluates the performance of lawmakers quantitatively and qualitatively.

[1128] Step 3: Compile the evaluation results into a "report card for the legislator."

[1129] Step 4: Publish the "Lawmakers' Report Cards."

[1130] Step 5: The emotion engine recognizes the user's emotion.

[1131] "Example 2"

[1132] Step 1: The user looks at the legislator's report card.

[1133] Step 2: The emotion engine captures the user's reaction via camera and microphone.

[1134] Step 3: The emotion engine analyzes the user's emotions from their facial expressions and tone of voice.

[1135] Step 4: Reflect the analyzed emotions when creating the next evaluation results.

[1136] "Example 3"

[1137] Step 1: Users comment on the results of the lawmaker's evaluation on social media.

[1138] Step 2: The sentiment engine analyzes the user's sentiment from the text.

[1139] Step 3: Reflect the analyzed information in the evaluation results.

[1140] Example 1

[1141] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1142] Conventional lawmaker evaluation systems mainly evaluated lawmakers based on quantitative data such as their speech content and attendance rate, and were unable to fully consider the emotions and qualitative aspects of their speech. Furthermore, when users looked at the evaluation results to decide who to vote for, they were not provided with a comprehensive evaluation that included emotional factors, resulting in a lack of information to support users' decision-making.

[1143] 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.

[1144] In this invention, the server includes means for acquiring published minutes data, means for preprocessing the acquired data, means for analyzing the preprocessed data using natural language processing technology, means for extracting information such as the content of remarks, frequency of remarks, and attendance rate from the analysis results, means for quantitatively and qualitatively evaluating the performance of the legislator based on the extracted information, means for creating and publishing the evaluation results as a "legislator report card," and an emotion engine that recognizes the user's emotions. This enables comprehensive evaluations that take into account not only the content of legislators' remarks and attendance rate, but also the emotions and qualitative aspects contained in the remarks. Furthermore, when users refer to the evaluation results to decide who to vote for, a comprehensive evaluation that includes emotional elements is provided, thereby enriching the information that supports the user's decision-making.

[1145] "Published minutes data" refers to record data of parliaments and meetings that are made public by governments and public institutions.

[1146] "Preprocessing" refers to processes such as data cleaning, tokenization, and removal of stop words that are carried out to prepare the acquired text data in a form that is easier to analyze.

[1147] "Natural language processing technology" is a technology that enables computers to understand and analyze human language, and is used for analyzing text and extracting information.

[1148] "Content of remarks" refers to the specific content and themes of remarks made by members of parliament in parliament or at meetings.

[1149] "Speech frequency" refers to the number of statements made by a member of parliament within a certain period of time.

[1150] "Attendance rate" is the percentage of times a member of parliament attended parliament or meetings.

[1151] "Quantitative and qualitative evaluation" means evaluation using both numerical data (quantitative evaluation) and qualitative aspects (qualitative evaluation).

[1152] A "legislator report card" is a report summarizing the results of an evaluation of a member of parliament's work, and is made public.

[1153] An "emotion engine" is a technology that analyzes emotions from text data and identifies emotions such as positive, negative, and neutral.

[1154] "Users" are general users who use the system to view the evaluation results of lawmakers.

[1155] This invention is a system that acquires publicly available minutes data and evaluates the performance of legislators using natural language processing technology and an emotion engine. A specific embodiment of this system is described below.

[1156] The server first obtains the publicly available meeting minutes data from the Internet. Specifically, it downloads the latest meeting minutes data using open data APIs provided by governments and public institutions. For example, it accesses the API using the Python requests library and obtains the data in JSON format.

[1157] The server then preprocesses the data, including cleaning the text, tokenizing it, and removing stop words. This is done using Python's pandas library and regular expressions (re library). For example, it removes unnecessary HTML tags and special characters, and splits the text into words using NLTK's word_tokenize function.

[1158] The preprocessed data is then analyzed using natural language processing technology. The server uses natural language processing libraries such as NLTK and spaCy to extract information such as the content of the legislator's remarks, their frequency, and their attendance rate. For example, topic modeling technology (e.g., LDA) is used to classify the content of the remarks, and legislators' remarks are categorized into topics such as "economy," "education," and "diplomacy."

[1159] The server then uses an emotion engine to analyze the emotions contained in the legislator's speech. Specifically, it uses the IBM Watson Tone Analyzer API to classify speech as positive, negative, or neutral.

[1160] Based on these analysis results, the server evaluates the performance of the legislators quantitatively and qualitatively. The evaluation is carried out using indicators such as frequency of speech, attendance rate, and emotional balance. The evaluation results are generated as a "legislator report card" and made public through a web interface.

[1161] A user can check the evaluation of a specific legislator using a web browser. For example, if a user requests "Show me the report card of legislator A," the server retrieves the legislator A's evaluation data from the database and displays it on the web page.

[1162] Examples of prompts to input to a generative AI model include:

[1163] "Analyze the text data of the Diet minutes and extract the content of statements made by Diet members, their frequency of statements, and their attendance rates. Furthermore, use an emotion engine to analyze the emotions contained in the statements and evaluate the performance of Diet members quantitatively and qualitatively."

[1164] In this way, the system analyzes the minutes data and performs a series of processes to evaluate the performance of the members of parliament.

[1165] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1166] Step 1: Data Acquisition

[1167] The server retrieves publicly available meeting minutes data from the Internet. Specifically, it downloads the latest meeting minutes data using open data APIs provided by governments and public institutions. It accesses the API using the Python requests library and retrieves the data in JSON format. The input is the API endpoint URL, and the output is the retrieved meeting minutes data in JSON format.

[1168] Step 2: Data Preprocessing

[1169] The server preprocesses the acquired JSON-formatted meeting minutes data. Specifically, it converts the data into a data frame using Python's pandas library and removes unnecessary HTML tags and special characters using regular expressions (the re library). Next, it splits the text into words using NLTK's word_tokenize function and removes unnecessary words by referring to a stop word list. The input is the acquired JSON-formatted meeting minutes data, and the output is the preprocessed text data.

[1170] Step 3: Natural Language Processing Analysis

[1171] The server analyzes the preprocessed text data using natural language processing technology. Specifically, it uses natural language processing libraries such as NLTK and spaCy to extract information such as the content of the legislator's remarks, frequency of remarks, and attendance rate. For example, it uses topic modeling technology (e.g., LDA) to classify the content of remarks, categorizing legislator remarks into topics such as "economy," "education," and "diplomacy." The input is the preprocessed text data, and the output is the analysis results, such as the content of remarks, frequency of remarks, and attendance rate.

[1172] Step 4: Sentiment Analysis

[1173] The server uses an emotion engine to analyze the emotions contained in the legislator's statements. Specifically, it uses the IBM Watson Tone Analyzer API to classify statements as positive, negative, or neutral. The input is preprocessed text data, and the output is the result of the emotion analysis.

[1174] Step 5: Generate evaluation results

[1175] The server quantitatively and qualitatively evaluates the performance of legislators based on the results of natural language processing and sentiment analysis. Evaluations are made using indicators such as speech frequency, attendance rate, and emotional balance. The evaluation results are generated as a "legislator's report card." The inputs are the analysis results of speech content, speech frequency, attendance rate, and sentiment analysis results, and the output is the evaluation results in the form of a "legislator's report card."

[1176] Step 6: Publication of evaluation results

[1177] The server makes the generated "report card of the legislator" available to the public through a web interface. Users can check the evaluation of a specific legislator using a web browser. For example, if a user requests "Display the report card of legislator A," the server retrieves the evaluation data of legislator A from the database and displays it on the web page. The input is the user's request, and the output is the evaluation result displayed on the web page.

[1178] (Application example 1)

[1179] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1180] Conventional lawmaker evaluation systems evaluated lawmakers based on quantitative data such as their speech content and attendance rate, but lacked the ability to provide personalized information that reflected the user's emotions and interests. They also lacked the ability to recommend lawmakers and topics that users might be interested in, making it difficult for users to efficiently obtain the information they needed. This limited their usefulness as reference information when deciding who to vote for.

[1181] 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.

[1182] In this invention, the server includes: means for AI to learn using published Diet minutes and the like as training data; means for the trained AI to quantitatively and qualitatively evaluate the performance of Diet members; means for creating and publishing the evaluation results as "Diet member report cards"; means including an emotion engine that recognizes the user's emotions; means for recommending Diet members and topics that the user may be interested in; and means for notifying the user of the latest statements and activities of Diet members they follow. This makes it possible to provide personalized information that reflects the user's emotions and interests, improving the usefulness of the information as reference when the user is deciding who to vote for.

[1183] "Published Diet minutes" are documents that record discussions and statements made in the Diet and are available to the public.

[1184] "Teacher data" is a dataset used for learning by a machine learning model, and is data that has been assigned a correct answer label.

[1185] "AI" is an abbreviation for artificial intelligence, a technology that allows computers to imitate human intelligence to learn and reason.

[1186] "Quantitative and qualitative evaluation" means evaluation based on numerical data (quantitative evaluation) and evaluation based on qualitative data (qualitative evaluation).

[1187] A "Member of Parliament Report Card" is a report summarizing the results of an evaluation based on the member's activities, statements, attendance rate, etc.

[1188] An "emotion engine" is a software or hardware system for recognizing and analyzing a user's emotions.

[1189] "Recommending politicians and topics that may be of interest to the user" means presenting relevant politicians and topics based on the user's interests and emotions.

[1190] "Notifying users of the latest statements and activities of politicians they follow" means providing users with real-time information on politicians they are interested in.

[1191] A system for carrying out this invention is configured as follows: The server includes means for AI to learn using published Diet minutes and the like as training data, means for the trained AI to quantitatively and qualitatively evaluate the performance of Diet members, means for creating and publishing the evaluation results as "Diet member report cards," means including an emotion engine that recognizes the user's emotions, means for recommending Diet members and topics that may interest the user, and means for notifying the user of the latest statements and activities of Diet members that the user follows.

[1192] Program processing explanation

[1193] The server first obtains the text data of the published Diet minutes and analyzes it using natural language processing technology. This analysis uses Hugging Face's Transformers library. As a result of the analysis, information such as the content of the Diet members' remarks, frequency of remarks, and attendance rate is extracted, and based on this, the performance of the Diet members is evaluated quantitatively and qualitatively. The evaluation results are compiled as a "Diet Member Report Card" and made publicly available for users to refer to.

[1194] Furthermore, the server uses an emotion engine to recognize the user's emotions. This emotion engine analyzes the user's input and behavioral data and recommends politicians and topics that the user may be interested in. This allows the user to efficiently obtain information based on their interests.

[1195] Users will be notified in real time of the latest statements and activities of the politicians they follow, allowing them to quickly keep up with the latest information on politicians they are interested in.

[1196] Specific examples

[1197] For example, if a user is interested in "environmental issues," the server will prioritize analyzing statements made by lawmakers on environmental issues and classify them into positive and negative. It will also recommend related lawmakers and topics based on the user's sentiment, allowing users to efficiently obtain information that matches their interests.

[1198] Prompt Sentence Examples

[1199] "Analyze the statements made by lawmakers on environmental issues and classify them into positive and negative. Also, recommend lawmakers that users might find interesting."

[1200] In this way, it becomes possible to provide personalized information that reflects the user's emotions and interests, improving the usefulness of the information as reference when the user decides who to vote for.

[1201] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1202] Step 1:

[1203] The server obtains the published text data of the Diet minutes. The input is the text data of the Diet minutes, and the output is text data in an analyzable format. Specifically, the server downloads the Diet minutes from a specified database or website and converts them into text format.

[1204] Step 2:

[1205] The server analyzes the acquired text data using natural language processing technology. The input is the text data acquired in step 1, and the output is information such as the content of each member's remarks, frequency of remarks, and attendance rate. Specifically, the server uses Hugging Face's Transformers library to analyze the text data and extract the content and frequency of each member's remarks.

[1206] Step 3:

[1207] The server evaluates the performance of the legislators quantitatively and qualitatively based on the analysis results. The input is the analysis results obtained in step 2, and the output is the legislator's evaluation score. Specifically, the server calculates a score based on the positive / negative evaluation of the content of the speech and attendance rate, and evaluates each legislator.

[1208] Step 4:

[1209] The server creates and publishes the evaluation results as "report cards for legislators." The input is the evaluation scores obtained in step 3, and the output is report card-style data that users can refer to. Specifically, the server organizes the evaluation scores and converts them into a format that can be viewed on a web page or application.

[1210] Step 5:

[1211] The server recognizes the user's emotions using an emotion engine. The input is the user's input data and behavioral data, and the output is the user's emotional state. Specifically, the server analyzes the user's text input and click history and uses the emotion engine to infer the user's emotional state.

[1212] Step 6:

[1213] The server recommends politicians and topics that the user is likely to be interested in. The input is the user's emotional state obtained in step 5 and the report card data obtained in step 4, and the output is the recommended politicians and topics. Specifically, the server selects relevant politicians and topics based on the user's emotional state and generates a recommendation list.

[1214] Step 7:

[1215] The server notifies the user of the latest statements and activities of the legislators they follow. The input is a list of legislators the user follows and the latest legislator activity data, and the output is a notification message. Specifically, the server references the follow list and notifies the user of relevant updates in real time.

[1216] Example 2

[1217] Next, a description will be given of Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1218] Conventional lawmaker evaluation systems typically base their evaluations on quantitative data such as the activities, statements, and attendance rate of lawmakers. However, these evaluations do not take into account the emotions and reactions of users, so the evaluation results do not necessarily reflect the opinions of voters. In addition, there are limited means for publishing the evaluation results, and they are often not provided in a format that is easily accessible to users. This creates the problem that the results are not fully utilized as reference information when voters decide who to vote for.

[1219] 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.

[1220] In this invention, the server includes means for an artificial intelligence to learn using published minutes and the like as training data, means for the trained artificial intelligence to quantitatively and qualitatively evaluate the performance of legislators, means for saving the evaluation results in a database, means for generating a report card in HTML format based on the saved evaluation results, means for publishing the generated report card on a website, means for recognizing the user's emotions, and means for reflecting the recognized emotions in the next evaluation results. This makes it possible to generate evaluation results that reflect the user's emotions and publish them in a form that users can easily access.

[1221] "Published minutes, etc." are documents and data that record the contents of assemblies and meetings and are made publicly available.

[1222] "Artificial intelligence" is a technology that mimics human intelligence through machine learning and data analysis.

[1223] "Quantitative and qualitative evaluation" means evaluating the value and performance of an object using numerical and qualitative data.

[1224] A "database" is a system for efficiently storing, managing, and retrieving data.

[1225] An "HTML report card" is a document in which the evaluation results are written using HTML (HyperText Markup Language) and formatted so that they can be displayed in a web browser.

[1226] A "website" is a collection of information that is publicly available on the Internet and accessible to users.

[1227] "Recognizing a user's emotions" means using sensors such as a camera and microphone to analyze the user's facial expressions and tone of voice and identify their emotions.

[1228] "Reflecting in the next evaluation result" means incorporating the recognized user emotion data into the next evaluation process to influence the evaluation result.

[1229] The present invention relates to a system for generating and publishing evaluation results of legislators. A specific embodiment of this system will be described below.

[1230] System configuration

[1231] Hardware

[1232] Server: manages the database, generates evaluation results, and hosts the website.

[1233] Terminal: The device used by the user to access the site (PC, smartphone, tablet, etc.).

[1234] Camera and microphone: sensors to recognize user emotions.

[1235] software

[1236] Database management system: MySQL, PostgreSQL, etc.

[1237] Web frameworks: Python, such as Django or Flask.

[1238] Emotion recognition libraries: OpenCV, TensorFlow, etc.

[1239] Program processing

[1240] Evaluation results stored in a database

[1241] The server receives the evaluation results of the legislator and stores them in a database such as MySQL or PostgreSQL. For example, if the evaluation results of legislator A are "Policy implementation ability: 8, Communication ability: 7, Reliability: 9," this is stored in the database.

[1242] HTML report card generation

[1243] The server retrieves the results from the database and generates an HTML report card using Django or Flask. The generated HTML file is published through a web server such as Apache or Nginx. Users can view the report card through a browser.

[1244] User Emotion Recognition

[1245] When a user uses a device to view a legislator's report card, the device's camera and microphone are used to capture the user's facial expressions and tone of voice in real time. For example, if the user smiles, the device uses OpenCV to recognize the smile and TensorFlow to analyze the tone of voice.

[1246] Reflection of evaluation results by emotion engine

[1247] The device sends the analyzed emotion data to the server. The server stores this data in a database and reflects it when creating the next evaluation result. For example, if the data "User ID: 123, Congressman ID: A, Emotion: Positive" is sent, it will have a positive effect on the next evaluation result.

[1248] Specific examples

[1249] For example, if a user smiles while looking at Congressman A's report card, the device will recognize the smile using OpenCV and send it to the server as a positive emotion. The server will then reflect this emotion data in the next evaluation result, slightly increasing Congressman A's evaluation.

[1250] Prompt Sentence Examples

[1251] Examples of prompts to be input to a generative AI model include:

[1252] "Please analyze the user's emotions after seeing Congressman A's report card and reflect them in the next evaluation results. If the user smiles, please consider that emotion as positive and add it to the evaluation."

[1253] In this way, the system can reflect the user's feelings in the evaluation results and create a more accurate "report card for the legislator."

[1254] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1255] Step 1: Save the evaluation results in a database

[1256] The server receives the evaluation results of the legislator. As input, it receives the legislator's evaluation result data (e.g., "Policy implementation ability: 8, Communication ability: 7, Trustworthiness: 9"). The server stores this data in a database such as MySQL or PostgreSQL. Specifically, it executes an SQL query to insert the evaluation results into the "evaluation_results" table. As output, it obtains the evaluation results stored in the database.

[1257] Step 2: Generate an HTML report card

[1258] The server retrieves the evaluation results from the database. It executes a query (e.g., "SELECT FROM evaluation_results WHERE legislator ID = 'A'") with the evaluation results retrieved from the database as input. The server uses Python's Django or Flask to embed the evaluation results into an HTML template. Specifically, it uses a template engine to convert the evaluation results into HTML format. The generated HTML file is obtained as output. The server publishes this HTML file through a web server such as Apache or Nginx.

[1259] Step 3: Recognizing user emotions

[1260] A user uses a device to access a website and view a representative's report card. As input, the user's facial expressions and tone of voice are acquired in real time via a camera and microphone. The device then analyzes the user's emotions using libraries such as OpenCV and TensorFlow. Specifically, it uses a facial recognition algorithm to identify emotions such as smiling or anger, and a voice analysis algorithm to analyze the tone of voice. The analyzed emotional data (e.g., "Emotion: Positive") is obtained as output.

[1261] Step 4: Reflecting the evaluation results using the emotion engine

[1262] The device sends the analyzed emotion data to the server. As input, it sends emotion data (e.g., "User ID: 123, Congressman ID: A, Emotion: Positive"). The server stores this data in a database. Specifically, it executes an SQL query to insert the emotion data into the "user_emotions" table. The next time it creates an evaluation result, the server takes the stored emotion data into account and reflects it in the evaluation result. As output, it obtains a new evaluation result that reflects the emotion data.

[1263] In this way, the system can reflect the user's feelings in the evaluation results and create a more accurate "report card for the legislator."

[1264] (Application example 2)

[1265] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1266] Conventional lawmaker evaluation systems evaluate lawmakers based on quantitative and qualitative data such as their activities, statements, and attendance rates, but do not take into account the user's emotions and reactions. As a result, the evaluation results may not reflect the user's actual emotions and opinions, and users are not provided with sufficient useful information. Furthermore, in the shopping experience at physical stores, personalized product recommendations that take the user's emotions into account are not made, making it difficult to improve user satisfaction.

[1267] 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.

[1268] In this invention, the server includes a means for AI to learn using published Diet minutes and other data as training data, a means for the trained AI to quantitatively and qualitatively evaluate the performance of Diet members, a means for creating and publishing the evaluation results as "Diet member report cards," a means for recognizing users' emotions and reflecting the results in the evaluation results, and a means for recognizing users' emotions when viewing products and making personalized product recommendations based on that information. This makes it possible to provide Diet member evaluation results that reflect users' emotions and to make personalized product recommendations in physical stores.

[1269] "Published Diet minutes" are documents that record discussions and statements made in the Diet and are available to the public.

[1270] "Training data" is a set of input data and its correct labels used for learning a machine learning model.

[1271] "AI" is an abbreviation for artificial intelligence, a technology that allows computers to imitate human intelligence to learn and reason.

[1272] "Quantitative and qualitative evaluation" means evaluating an object using numerical and qualitative data.

[1273] A "legislator's report card" is a report summarizing the activities of a member of parliament and the results of their evaluation.

[1274] "Recognizing the user's emotions" means analyzing the user's emotions from their facial expressions, tone of voice, etc.

[1275] "Personalized product recommendation" refers to recommending products based on a user's individual preferences and emotions.

[1276] A "system" is a collection of devices and software that combine multiple elements to achieve a specific function.

[1277] A system for implementing the present invention is configured as follows.

[1278] System configuration

[1279] The system consists of a server, a user terminal (e.g., smart glasses), and software components such as an emotion recognition engine, a product recommendation engine, and a display module.

[1280] server

[1281] The server has a means for the AI ​​to learn from publicly available Diet minutes and other training data. The trained AI then quantitatively and qualitatively evaluates the performance of legislators, creating and publishing the results as "legislator report cards." It also has a means for recognizing users' emotions and reflecting those results in the evaluation results.

[1282] User terminal

[1283] The user terminal is a wearable device such as smart glasses, equipped with a camera and a display. It has a means of recognizing the user's emotions when viewing products and making personalized product recommendations based on that information.

[1284] Hardware and software used

[1285] Hardware: Smart glasses (camera, display), computer (built into the smart glasses)

[1286] Software: OpenCV (camera image acquisition and processing), EmotionRecognizer (emotion recognition engine), ProductRecommender (product recommendation engine), DisplayModule (display module)

[1287] Data processing and calculation

[1288] server

[1289] The server uses the published Diet minutes as training data for the AI ​​to learn from. After learning, the AI ​​evaluates the lawmakers based on data such as their activities, statements, and attendance rates. The evaluation results are generated as "lawmaker report cards" and published on a website. The server also recognizes the user's emotions and reflects the results in the next evaluation.

[1290] User terminal

[1291] The user device captures the user's facial expressions in real time through the smart glasses' camera and analyzes their emotions using EmotionRecognizer. The analyzed emotional data is sent to ProductRecommender, which then recommends products based on the user's emotions. Recommended product information is then displayed on the smart glasses' display.

[1292] Specific examples

[1293] If a user wears smart glasses in a physical store and expresses surprise when looking at a particular product, related new products and promotional information will be displayed based on that emotion.

[1294] Prompt Sentence Examples

[1295] "Develop an application for smart glasses that recognizes the emotions users feel when viewing products and provides personalized product recommendations based on those emotions. Include a feature to display new products and promotional information if the user is surprised."

[1296] The above is an embodiment of the present invention.

[1297] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1298] Step 1:

[1299] The server uses the published Diet minutes as training data to train the AI. The input is the text data of the Diet minutes, and the output is the trained AI model. Specifically, it analyzes the text data using natural language processing technology and extracts the statements and activities of Diet members as features.

[1300] Step 2:

[1301] The server uses the trained AI to quantitatively and qualitatively evaluate the performance of the legislators. The input is the legislator's activity data, and the output is the evaluation results. Specifically, the AI ​​model calculates a score based on data such as the legislator's remarks and attendance rate, and generates the evaluation results.

[1302] Step 3:

[1303] The server creates the evaluation results as "report cards for legislators" and publishes them on a website. The input is the evaluation results, and the output is the report card in HTML format. Specifically, the server saves the evaluation results in a database and generates an HTML file to display as a web page.

[1304] Step 4:

[1305] The user device captures the user's facial expressions in real time through the smart glasses' camera. The input is the camera image, and the output is the user's facial expression data. Specifically, the image captured by the camera is processed frame by frame, and facial expressions are analyzed using a face detection algorithm.

[1306] Step 5:

[1307] The user device analyzes the user's emotions using EmotionRecognizer. The input is the user's facial expression data, and the output is emotion data. Specifically, the facial expression data is input into an emotion recognition engine, which classifies emotions such as joy, anger, and surprise.

[1308] Step 6:

[1309] The user device sends the analyzed emotional data to the ProductRecommender, which then makes personalized product recommendations. The input is emotional data, and the output is a list of recommended products. Specifically, the system infers the user's preferences and interests based on the emotional data, and recommends related products.

[1310] Step 7:

[1311] The user terminal displays the recommended product information on the display of the smart glasses. The input is the recommended product list, and the output is the product information displayed on the display. Specifically, the recommended product list is input to the display module and displayed in a form that can be visually confirmed by the user.

[1312] The above are the specific processing steps for carrying out the present invention.

[1313] Example 3

[1314] Next, a description will be given of Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1315] Conventional legislator evaluation systems mainly evaluate legislators based on quantitative data such as their activities and attendance rates, and are unable to reflect the emotions and opinions of users. As a result, when users use the evaluation results to decide who to vote for, their individual emotions and opinions are not reflected, reducing the reliability and usefulness of the evaluation results. Furthermore, the way the evaluation results are displayed can sometimes be difficult to understand visually, making it difficult for users to understand.

[1316] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[1317] In this invention, the server includes: means for an artificial intelligence to learn using publicly available parliamentary records and other training data; means for the trained artificial intelligence to quantitatively and qualitatively evaluate the performance of legislators; means for creating and publishing the evaluation results as a "legislator report card"; means for analyzing facial expressions, voice, and text input to recognize the user's emotions; means for reflecting the recognized emotional data in the evaluation results; and means for displaying the evaluation results in a form that can be referenced by the user. This makes it possible to provide evaluation results that reflect the user's emotions and opinions, improving the reliability and usefulness of users when using the evaluation results to decide who to vote for. Furthermore, displaying the evaluation results in a visually easy-to-understand format makes them easy for users to understand.

[1318] "Public parliamentary records" are documents containing data such as members' activities, statements, and attendance rates that can be obtained from official parliamentary websites and other public sources.

[1319] "Artificial intelligence" refers to computer systems that use techniques such as machine learning and deep learning to learn patterns from large amounts of data and automatically perform specific tasks.

[1320] "Quantitative and qualitative evaluation" means to comprehensively evaluate the work of legislators by combining evaluation based on numerical data (quantitative evaluation) with evaluation based on qualitative data (qualitative evaluation).

[1321] A "legislator report card" is a report summarizing the evaluation results of a member of parliament's activities, attendance rate, comments, etc., and is made publicly available for users to refer to.

[1322] "Recognizing user emotions" means identifying the emotions a user is feeling by analyzing their facial expressions, voice, text input, etc.

[1323] "Reflecting emotional data in the evaluation results" means adding or subtracting from the lawmaker's evaluation score based on the user's emotion recognition results, and reflecting this in the final evaluation results.

[1324] "Displaying the evaluation results in a form that users can refer to" means displaying the evaluation results on a website in a visually easy-to-understand format, such as a graph or table, so that users can easily understand them.

[1325] MODE FOR CARRYING OUT THE INVENTION

[1326] This invention is a system that provides voters with evaluation results for each legislator as reference information when deciding who to vote for. Specifically, voters refer to each legislator's report card on a website and decide who to vote for based on the evaluation results. It also uses an emotion engine to recognize the user's emotions and reflect them in the evaluation results.

[1327] Hardware and software used

[1328] 1. Server

[1329] Hardware: High-performance server machine

[1330] Software: Python, BeautifulSoup, pandas, OpenCV

[1331] 2. Terminal

[1332] Hardware: Your PC or smartphone

[1333] Software: Web browser, JavaScript, D3.js

[1334] 3. Users

[1335] Hardware: Webcam, microphone

[1336] Software: Web browser

[1337] Data processing and calculation

[1338] 1. Data Collection

[1339] The server uses Python's BeautifulSoup library to scrape parliamentary attendance data from official parliamentary websites, for example by visiting specific URLs and extracting attendance information from specific HTML tags.

[1340] 2. Data analysis

[1341] The server uses the Python pandas library to analyze the data and calculate an evaluation score for each legislator. For example, it implements an algorithm that gives a higher evaluation to legislators with an attendance rate of 90% or higher.

[1342] 3. Display of evaluation results

[1343] The device uses the JavaScript library D3.js to display the evaluation results in a graphical format, for example, displaying each legislator's evaluation score in a bar graph, allowing users to easily compare them.

[1344] 4. Emotion recognition

[1345] The emotion engine uses OpenCV to analyze the user's facial expressions and detect positive emotions. For example, if the user smiles through the webcam, it will recognize that smile as a positive emotion.

[1346] 5. Reflecting Emotional Data

[1347] If a user comments, "This legislator is doing a great job," the server adds that positive sentiment to the rating score. For example, the server implements an algorithm that gives additional points to legislators with many positive comments.

[1348] Specific examples

[1349] When a user accesses the website and views a member of parliament's report card, they can input a prompt sentence like the one below into the generative AI model to generate a sentence explaining the system's overview and processing procedures in natural language.

[1350] "Please explain the system that decides who to vote for based on the results of an evaluation of a member of parliament. Specifically, the system refers to each member of parliament's report card on a website and decides who to vote for based on the evaluation results. It also uses an emotion engine to recognize the user's emotions and reflect them in the evaluation results."

[1351] By inputting this prompt sentence into the generative AI model, a sentence explaining the system overview and processing procedure in natural language can be generated. The flow of the specific processing in Example 3 will be described with reference to FIG.

[1352] Program processing flow

[1353] Step 1: Data collection

[1354] Input: Official Congressional website URL

[1355] Output: Data on attendance rate and speech content of each member of parliament

[1356] Specific behavior:

[1357] The server uses Python's BeautifulSoup library to scrape the attendance data of members from the official parliamentary website. Specifically, it accesses specific URLs and extracts attendance information from specific HTML tags. For example, it parses a table listing each member's attendance rate to obtain the required data.

[1358] Step 2: Data analysis

[1359] Input: Data on attendance and speech content of each member of parliament collected in Step 1

[1360] Output: Evaluation score for each legislator

[1361] Specific behavior:

[1362] The server analyzes the data using Python's pandas library and calculates an evaluation score for each legislator. Specifically, it implements an algorithm that gives higher evaluations to legislators with attendance rates of 90% or higher. For example, it calculates an overall evaluation score based on attendance rate, number of comments, and quality of comments.

[1363] Step 3: View the evaluation results

[1364] Input: Evaluation scores for each legislator calculated in step 2

[1365] Output: Evaluation results displayed on the website

[1366] Specific behavior:

[1367] The device uses the JavaScript library D3.js to display the evaluation results in a graph format. Specifically, it displays each legislator's evaluation score in a bar graph, allowing users to easily compare them. For example, it generates a graph with the evaluation score on the vertical axis and the legislator's name on the horizontal axis.

[1368] Step 4: Emotion Recognition

[1369] Input: User facial expressions, voice, and text input

[1370] Output: User sentiment data

[1371] Specific behavior:

[1372] The emotion engine uses OpenCV to analyze the user's facial expressions and detect positive emotions. Specifically, if the user smiles through the webcam, it recognizes the smile as a positive emotion. For example, it analyzes facial feature points to detect smile patterns.

[1373] Step 5: Reflecting emotional data

[1374] Input: User emotion data obtained in step 4

[1375] Output: Updated rating score

[1376] Specific behavior:

[1377] When a user comments, "This legislator is doing a great job," the server adds that positive sentiment to the evaluation score. Specifically, the server implements an algorithm that gives additional points to legislators with many positive comments. For example, the server reflects the positive sentiment data in the evaluation score and updates the final evaluation result.

[1378] (Application example 3)

[1379] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1380] In conventional lawmaker evaluation systems, evaluation results are provided unilaterally and are unable to reflect the user's feelings or opinions. In addition, there are limited ways to check the evaluation results in real time, which means that users cannot fully utilize the information when deciding who to vote for.

[1381] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[1382] In this invention, the server includes: means for AI to learn using published Diet minutes and the like as training data; means for the trained AI to quantitatively and qualitatively evaluate the performance of Diet members; means for creating and publishing the evaluation results as "Diet member report cards"; means for analyzing the user's facial expressions, voice, text input, etc. using an emotion engine to recognize the user's emotions; means for reflecting the user's emotional information in the evaluation results; and means for the user to check the evaluation results in real time through an application installed on a smartphone. This makes it possible to provide evaluation results that reflect the user's emotions and opinions, and for the user to use them in real time as reference information when deciding who to vote for.

[1383] "Published Diet minutes" are documents that record the discussions and decisions made in the Diet and are available to the public.

[1384] "Training data" is a set of input data and its correct labels used for learning a machine learning model.

[1385] "AI" is an abbreviation for artificial intelligence, a technology that allows computers to imitate human intelligence to learn and reason.

[1386] "The work of a member of parliament" refers to the activities, statements, and attendance of members of parliament in the Diet and other official duties.

[1387] "Quantitative and qualitative evaluation" means that evaluation is conducted using both numerical and qualitative data.

[1388] A "Member of Parliament Report Card" is a report that evaluates a member of parliament's activities, comments, attendance, etc. and summarizes the results.

[1389] An "emotion engine" is software or algorithms that recognize and analyze user emotions.

[1390] "Facial expressions" refer to the movements and expressions of a user's face, and are one way of expressing emotions.

[1391] "Voice" refers to the voice or spoken words produced by the user, and is one of the means of expressing emotions.

[1392] "Text input" refers to textual information entered by a user using a keyboard or touchscreen.

[1393] A "smartphone" is a type of mobile phone and a multi-function device that can connect to the Internet and use applications.

[1394] "Real-time" refers to data and information being processed and provided to users immediately.

[1395] A system for implementing this invention is configured as follows: The server has a means for AI to learn from published Diet records and other data as training data. Specifically, the server collects Diet records and supplies them to a machine learning model as training data. This allows the AI ​​to learn about the activities and statements of Diet members and evaluate their performance quantitatively and qualitatively.

[1396] Next, the server has a means for the trained AI to quantitatively and qualitatively evaluate the performance of the legislators. The evaluation is calculated based on the legislators' activities, comments, attendance rate, etc. The evaluation results are compiled as "legislator report cards" and made public. This allows users to refer to the evaluation results and use them as reference information when deciding who to vote for.

[1397] Furthermore, the server has a means to analyze the user's facial expressions, voice, text input, etc. using an emotion engine to recognize the user's emotions. The emotion engine collects and analyzes the emotions expressed by the user regarding the results of the legislator's evaluation. For example, if a user comments, "This legislator is doing a great job," the text conveys the user's positive emotions. This information is also collected and analyzed by the emotion engine and reflected in the evaluation results.

[1398] The server also has a means to reflect the user's emotional information in the evaluation results, which allows the evaluation results to reflect the user's emotions and opinions, making it possible to provide more reliable information.

[1399] Finally, the server provides a means for users to check the evaluation results in real time through an application installed on their smartphones. Users can use their smartphones to check the evaluation results of lawmakers in real time and use this information as reference when deciding who to vote for.

[1400] As a concrete example, consider the case where a user opens the "Diet Member Evaluation Navigator" app on their smartphone to check the evaluation results of a specific legislator. When the user comments, "This legislator is doing a great job," and uploads a photo of their face, the app analyzes their emotions from the comment and photo and reflects this in the evaluation results. In this way, it is possible to provide evaluation results that reflect the user's emotions and opinions.

[1401] Examples of prompts for generative AI models include:

[1402] Open the "Assemblyman Evaluation Navigator" app and view the evaluation results for assemblyman ID "12345." Upload a user comment such as "This assemblyman is doing a great job," along with a photo of the person's face, and have the system analyze their emotions and reflect them in the evaluation results.

[1403] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[1404] Step 1:

[1405] The server collects published parliamentary records.

[1406] Input: Published Parliamentary Records

[1407] Output: Collected parliamentary records data

[1408] Specific operation: The server periodically crawls publicly available parliamentary minutes on the Internet and stores them in a database.

[1409] Step 2:

[1410] The server supplies the collected parliamentary minutes as training data to the machine learning model, allowing the AI ​​to learn.

[1411] Input: Collected parliamentary transcript data

[1412] Output: Trained AI model

[1413] Specific operation: The server preprocesses the collected parliamentary transcript data and feeds it to a machine learning algorithm to train a model.

[1414] Step 3:

[1415] The server uses a trained AI model to quantitatively and qualitatively evaluate the performance of lawmakers.

[1416] Input: Trained AI model, lawmaker activity data

[1417] Output: Evaluation results of legislators

[1418] Specific operation: The server inputs the activity data of the lawmakers into a trained AI model and generates evaluation results.

[1419] Step 4:

[1420] The server creates and publishes the evaluation results as "report cards for legislators."

[1421] Input: Evaluation results of the legislator

[1422] Output: "Congressman's report card"

[1423] Specific operation: The server formats the evaluation results and makes them available through a website or application.

[1424] Step 5:

[1425] The device uses an emotion engine to recognize the user's emotions by analyzing the user's facial expressions, voice, text input, etc.

[1426] Input: User facial expressions, voice, and text input

[1427] Output: User emotion data

[1428] Specific operation: The device uses a camera and microphone to capture the user's facial expressions and voice, receives text input, and analyzes it with an emotion engine.

[1429] Step 6:

[1430] The server reflects the user's emotional information in the evaluation results.

[1431] Input: User's emotional data, legislator evaluation results

[1432] Output: Evaluation results reflecting emotional information

[1433] Specific operation: The server integrates the user's emotional data into the evaluation results and generates evaluation results that reflect the emotional information.

[1434] Step 7:

[1435] The device allows users to check the evaluation results in real time through an application installed on their smartphone.

[1436] Input: Evaluation results that reflect emotional information

[1437] Output: Evaluation results for user review

[1438] Specific operation: The device displays the evaluation results through a smartphone app, allowing users to check them in real time.

[1439] 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.

[1440] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1441] Another example of generative AI is Gemini (internet search engine). <url: https: gemini.google.com ?hl="ja">) are listed.

[1442] 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.

[1443] [Third embodiment]

[1444] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1445] 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.

[1446] 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).

[1447] 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.

[1448] 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.

[1449] 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).

[1450] 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.

[1451] 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.

[1452] 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.

[1453] 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.

[1454] 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.

[1455] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[1456] "Example 1"

[1457] One embodiment of the present invention is a system in which AI learns from publicly available Diet minutes and other training data. Specifically, the system acquires text data from Diet minutes and analyzes it using natural language processing technology. The analysis extracts information such as the content and frequency of Diet members' remarks, attendance rate, etc., and uses this information to quantitatively and qualitatively evaluate the performance of Diet members.

[1458] "Example 2"

[1459] The evaluation results are compiled as "report cards for legislators" and made public. Specifically, the evaluation results for each legislator are stored in a database, and an HTML report card is generated based on the results. The generated report card is made public on a website so that voters can view it.

[1460] "Example 3"

[1461] Another embodiment of the present invention is a system that uses the evaluation results as reference information when voters decide who to vote for. Specifically, the system refers to each legislator's report card on a website and decides who to vote for based on the evaluation results. For example, it is possible to give priority to legislators with high attendance rates or high evaluations of the content of their statements.

[1462] The processing flow of each embodiment will be described below.

[1463] "Example 1"

[1464] Step 1: Obtain text data of the Diet minutes from the Internet.

[1465] Step 2: The acquired text data is analyzed using natural language processing technology. This analysis extracts information such as the content and frequency of the legislator's statements and attendance rate.

[1466] Step 3: Based on the extracted information, the AI ​​quantitatively and qualitatively evaluates the performance of the lawmakers.

[1467] "Example 2"

[1468] Step 1: Save the AI ​​evaluation results in a database.

[1469] Step 2: Generate an HTML report card based on the saved evaluation results.

[1470] Step 3: Publish the generated report card on a website so that voters can view it.

[1471] "Example 3"

[1472] Step 1: Voters view each representative's report card on the website.

[1473] Step 2: Voters decide who to vote for based on the evaluation results of the report card they have referenced. For example, they can prioritize lawmakers with high attendance rates or high evaluations of the content of their speeches.

[1474] Example 1

[1475] Next, a description will be given of Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1476] Conventional evaluation systems for legislators often require manual data collection and analysis to quantitatively and qualitatively evaluate legislators' performance, which is time-consuming and labor-intensive. Furthermore, evaluation results are not centrally managed and are not provided in an easily accessible format for users, making it difficult to use the evaluation results to decide who to vote for.

[1477] 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.

[1478] In this invention, the server includes: means for an artificial intelligence to learn using published minutes and the like as training data; means for the trained artificial intelligence to quantitatively and qualitatively evaluate the performance of legislators; means for creating and publishing the evaluation results as "legislator report cards"; means for acquiring text data of minutes via the Internet; means for preprocessing the acquired text data; means for analyzing the preprocessed data using natural language processing technology; means for evaluating the performance of legislators based on the analysis results; and means for outputting the evaluation results in a format that can be viewed by users. This makes it possible to efficiently and accurately evaluate the performance of legislators and provide the evaluation results in a format that users can easily refer to.

[1479] "Published minutes" are records of meetings that are made public by a government or public institution.

[1480] "Training data" is a dataset used for learning artificial intelligence.

[1481] "Artificial intelligence" is a technology that allows computer systems to behave in a way that mimics human intelligence.

[1482] "Quantitative evaluation" is a method of evaluating an object using numbers and statistics.

[1483] "Qualitative evaluation" is a method of evaluating an object by taking into account its qualitative aspects.

[1484] A "legislator report card" is a report summarizing the results of an evaluation of a member of parliament's work.

[1485] The "Internet" is a global communications network that interconnects computer networks around the world.

[1486] "Text data" is data that includes character information.

[1487] "Preprocessing" is the process of formatting and processing data before data analysis.

[1488] "Natural language processing technology" is a technology that uses computers to understand, analyze, and generate human language.

[1489] "Analysis" is the process of examining and evaluating data in detail.

[1490] "Evaluation results" are the results of evaluation obtained based on analysis.

[1491] "User" means a person or organization that uses this system.

[1492] "Format" is the specific way in which data or information is presented.

[1493] MODE FOR CARRYING OUT THE INVENTION

[1494] This invention is a system in which an artificial intelligence learns from publicly available minutes and other training data, and quantitatively and qualitatively evaluates the performance of legislators. A specific embodiment of this system will be described below.

[1495] 1. Program Generation

[1496] The server uses publicly available data such as meeting minutes as training data to build a generative AI model. This AI model is designed to analyze the text data in the minutes using natural language processing techniques. Specific software used includes Python's NLTK library and spaCy.

[1497] 2. Data Acquisition and Preprocessing

[1498] The terminal acquires the text data of the minutes via the Internet. The acquired data is sent to the server, which then performs the following preprocessing.

[1499] Text cleaning: Remove unnecessary symbols and spaces.

[1500] Tokenization: Splitting text into words and sentences.

[1501] Stop word removal: Remove meaningless words (e.g. "no", "wa", etc.).

[1502] 3. Data Analysis

[1503] The server analyzes the preprocessed data using natural language processing technology. Specifically, it performs the following analysis:

[1504] Speech analysis: Analyze the topic and sentiment of the speech.

[1505] Calculating speech frequency: Count the number of times each member of parliament speaks.

[1506] Calculating attendance rate: Extract the attendance rate of each member from the minutes.

[1507] 4. Generating ratings

[1508] The server evaluates the performance of the legislators based on the analysis results. The evaluation is performed as follows:

[1509] Quantitative evaluation: Quantify speaking frequency and attendance rate.

[1510] Qualitative evaluation: Evaluating the quality of what was said.

[1511] 5. Outputting the results

[1512] The server outputs the evaluation results in a format that can be viewed by the user, such as a web interface or a PDF report.

[1513] Specific examples

[1514] Specific use cases

[1515] If a user wants to evaluate the performance of a particular legislator, they can input a prompt like the following into the generative AI model:

[1516] Example prompt sentence:

[1517] "Based on the 2023 Diet minutes, please evaluate the content, frequency, and attendance rate of Congressman A."

[1518] When this prompt is entered, the following specific processing is performed:

[1519] 1. Data Acquisition:

[1520] The device will download the 2023 Diet minutes from the government's public API.

[1521] 2. Data preprocessing:

[1522] The server removes unnecessary symbols and spaces from the downloaded text data and performs tokenization.

[1523] The server removes stop words and converts the data into a format suitable for analysis.

[1524] 3. Data Analysis:

[1525] The server uses natural language processing technology to analyze the content of Congressman A's statements. For example, it detects that Congressman A frequently speaks about "economic policy."

[1526] The server calculates the frequency of Congressman A's statements and finds that the average number of times per month is 20.

[1527] The server calculates the attendance rate of Council Member A and extracts that it is 95%.

[1528] 4. Generate ratings:

[1529] The server quantitatively evaluates the performance of Councillor A based on frequency of speech and attendance rate.

[1530] The server qualitatively evaluates the performance of Congressman A based on the quality of his remarks.

[1531] 5. Result output:

[1532] The server generates the evaluation results as a PDF report that can be downloaded by the user.

[1533] In this way, the user can evaluate Congressman A's performance in detail.

[1534] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1535] Program processing flow

[1536] Step 1: Get the data

[1537] Input: The user enters a prompt to rate the performance of a particular legislator.

[1538] Operation: The device downloads text data of meeting minutes for a specified period from the government's public API via the Internet.

[1539] Output: Downloaded text data of the minutes.

[1540] Step 2: Preprocessing the data

[1541] Input: Text data of downloaded minutes.

[1542] Operation: The server performs the following preprocessing.

[1543] Text cleaning: Remove unnecessary symbols and spaces.

[1544] Tokenization: Splitting text into words and sentences.

[1545] Stop word removal: Remove meaningless words (e.g. "no", "wa", etc.).

[1546] Output: Preprocessed text data.

[1547] Step 3: Analyze the data

[1548] Input: Preprocessed text data.

[1549] Operation: The server performs the following analysis using natural language processing technology.

[1550] Speech analysis: Analyze the topic and sentiment of the speech.

[1551] Calculating speech frequency: Count the number of times each member of parliament speaks.

[1552] Calculating attendance rate: Extract the attendance rate of each member from the minutes.

[1553] Output: Analysis results (content of comments, frequency of comments, attendance rate).

[1554] Step 4: Generate ratings

[1555] Input: Analysis results (content of comments, frequency of comments, attendance rate).

[1556] How it works: The server evaluates the performance of the legislators based on the analysis results.

[1557] Quantitative evaluation: Quantify speaking frequency and attendance rate.

[1558] Qualitative evaluation: Evaluating the quality of what was said.

[1559] Output: Evaluation results (quantitative evaluation, qualitative evaluation).

[1560] Step 5: Output the results

[1561] Input: Evaluation results (quantitative evaluation, qualitative evaluation).

[1562] Behavior: The server outputs the evaluation results in a format that can be viewed by the user, such as a web interface or a PDF report.

[1563] Output: User-visible evaluation results.

[1564] Specific examples

[1565] Specific use cases

[1566] If a user wants to evaluate the performance of a particular legislator, they can input a prompt like the following into the generative AI model:

[1567] Example prompt sentence:

[1568] "Based on the 2023 Diet minutes, please evaluate the content, frequency, and attendance rate of Congressman A."

[1569] When this prompt is entered, the following specific processing is performed:

[1570] 1. Data Acquisition:

[1571] The device will download the 2023 Diet minutes from the government's public API.

[1572] 2. Data preprocessing:

[1573] The server removes unnecessary symbols and spaces from the downloaded text data and performs tokenization.

[1574] The server removes stop words and converts the data into a format suitable for analysis.

[1575] 3. Data Analysis:

[1576] The server uses natural language processing technology to analyze the content of Congressman A's statements. For example, it detects that Congressman A frequently speaks about "economic policy."

[1577] The server calculates the frequency of Congressman A's statements and finds that the average number of times per month is 20.

[1578] The server calculates the attendance rate of Council Member A and extracts that it is 95%.

[1579] 4. Generate ratings:

[1580] The server quantitatively evaluates the performance of Councillor A based on frequency of speech and attendance rate.

[1581] The server qualitatively evaluates the performance of Congressman A based on the quality of his remarks.

[1582] 5. Result output:

[1583] The server generates the evaluation results as a PDF report that can be downloaded by the user.

[1584] In this way, the user can evaluate Congressman A's performance in detail.

[1585] (Application example 1)

[1586] Next, a description will be given of Application Example 1 of Form Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1587] In traditional conferences and meetings, it was difficult to quantitatively and qualitatively evaluate participants' comments, frequency of comments, attendance rates, etc., resulting in problems that reduced the transparency and efficiency of the meetings. Furthermore, there was a lack of means to objectively evaluate participants' performance, making it impossible to use this information as reference for decision-making. This could result in a decline in the quality of meetings and a negative impact on the performance of the entire organization.

[1588] 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.

[1589] In this invention, the server includes a means for AI to learn using publicly available minutes and the like as training data, a means for the trained AI to quantitatively and qualitatively evaluate the performance of participants, a means for creating and publishing the evaluation results as a "participant report card," and a means for improving the transparency and efficiency of meetings. This makes it possible to objectively evaluate the content and frequency of participants' comments, attendance rates, etc., and improve the transparency and efficiency of meetings.

[1590] A "published minutes" is a document that records the contents of a conference or meeting and is available to the public.

[1591] "Training data" is a dataset used to train a machine learning model.

[1592] "AI" is an abbreviation for artificial intelligence, a system that mimics human intellectual tasks using technologies such as machine learning and natural language processing.

[1593] A "participant" is someone who attends a conference or meeting and expresses their opinions or comments.

[1594] "Quantitative evaluation" is a method of evaluation using numerical and statistical data.

[1595] "Qualitative evaluation" is a method of evaluation that takes into account qualitative aspects.

[1596] A "participant report card" is a document of evaluation results created based on the content and frequency of participants' comments, attendance rate, etc.

[1597] "Transparency" means that information is not hidden but is made public and available for anyone to see.

[1598] "Efficiency" is the state in which the time and effort required to achieve a goal is minimized.

[1599] A system for carrying out this invention is configured as follows: The server includes means for AI to learn using publicly available minutes and the like as training data, means for the trained AI to quantitatively and qualitatively evaluate the performance of participants, means for creating and publishing the evaluation results as "participant report cards," and means for improving the transparency and efficiency of meetings.

[1600] The server analyzes the minutes text using Python and the natural language processing library spaCy. Specifically, the server performs the following steps:

[1601] First, the server reads the minutes text from a file. Next, it analyzes the read text using spaCy to extract the content of each speaker's comments. The extracted content is categorized by speaker and the frequency of each comment is calculated. This allows for a quantitative evaluation of the content and frequency of each participant's comments.

[1602] The evaluation results are compiled as a "participant report card" and made publicly available for users to refer to, allowing them to use the information to improve the transparency and efficiency of their meetings.

[1603] As a concrete example, let's say we have the following meeting minutes:

[1604] Tanaka: Let's discuss a new project today.

[1605] Suzuki: Before that, let's check the minutes of the last meeting.

[1606] Tanaka: Understood. I will check the minutes.

[1607] An analysis of the minutes reveals the following results:

[1608] Tanaka: 2 comments

[1609] Suzuki: 1 statement

[1610] An example of a prompt sentence to input to the generative AI model is as follows:

[1611] Analyze the following meeting minutes and calculate the frequency of each speaker's comments.

[1612] Tanaka: Let's discuss a new project today.

[1613] Suzuki: Before that, let's check the minutes of the last meeting.

[1614] Tanaka: Understood. I will check the minutes.

[1615] In this way, the server can analyze the content and frequency of participants' comments and publish the evaluation results as a "participant report card," thereby improving the transparency and efficiency of meetings.

[1616] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1617] Step 1:

[1618] The server reads the published minutes text from a file. The input is the minutes text file, and the output is text data. This text data serves as the basis for subsequent natural language processing.

[1619] Step 2:

[1620] The server analyzes the loaded text data using spaCy, a natural language processing library. The input is the text data obtained in step 1, and the output is a parsed document object. This document object contains information such as sentence and word tokenization, part-of-speech tagging, and dependency analysis.

[1621] Step 3:

[1622] The server extracts the content of each speaker's comments from the analyzed document object. The input is the document object obtained in step 2, and the output is a list of comments categorized by speaker. Specifically, the speaker and the content of the comments are separated based on the ":" in the sentence, and the list is organized by speaker.

[1623] Step 4:

[1624] The server calculates the frequency of each speaker's comments. The input is the list of comments obtained in step 3, and the output is a dictionary of the frequency of each speaker's comments. Specifically, the server counts the number of comments made by each speaker and saves them in dictionary format.

[1625] Step 5:

[1626] The server creates a "participant report card" based on the dictionary of comment frequency. The input is the dictionary of comment frequency obtained in step 4, and the output is the evaluation results as a "participant report card." Specifically, the evaluation of each participant is quantitatively described based on the comment frequency and saved in document format.

[1627] Step 6:

[1628] The server publishes the created "Participant Report Cards" in a form that users can refer to. The input is the evaluation results obtained in step 5, and the output is a web page or dashboard that users can access. Specifically, the evaluation results are converted into HTML format and published on the web server.

[1629] In this way, the server analyzes the minutes text, evaluates the content and frequency of participants' comments, and publishes the results as a "participant report card," thereby improving the transparency and efficiency of meetings.

[1630] Example 2

[1631] Next, a description will be given of Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1632] In the conventional system for evaluating legislators, the evaluation results are not managed centrally, and the evaluations lack transparency and reliability. In addition, the evaluation results are not properly disclosed to the general public, which creates the problem that voters cannot use them as reference information when deciding who to vote for.

[1633] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes a means for an artificial intelligence to learn using published minutes and the like as training data, a means for the trained artificial intelligence to quantitatively and qualitatively evaluate the performance of the legislators, a means for saving the evaluation results in a database, a means for generating a report card in HTML format based on the saved evaluation results, and a means for publishing the generated report card on a website. This enables centralized management of evaluation results and ensures transparency, allowing voters to decide who to vote for based on reliable information.

[1634] "Published minutes, etc." are documents and data that record the content of statements and discussions made in official forums such as parliaments and conferences and are made publicly available.

[1635] "Artificial intelligence" is a technology that enables computer systems to mimic human intelligence and perform intellectual tasks such as learning, reasoning, and recognition.

[1636] The "means of learning" refers to the process by which AI analyzes data, such as publicly available minutes, to find patterns and rules.

[1637] "Quantitative and qualitative evaluation methods" refers to the process by which artificial intelligence evaluates the performance of legislators using numerical data (quantitative) and qualitative data (qualitative).

[1638] "Evaluation results" are data obtained as a result of AI evaluating the performance of lawmakers, and include the activities, statements, attendance rate, etc. of lawmakers.

[1639] A "database" is a digital information management system for efficiently storing, managing, and searching evaluation results.

[1640] An "HTML report card" is a web page document created based on the evaluation results to visually display the evaluation of a member of parliament.

[1641] "Publication on a website" is the process by which the generated HTML report card is made accessible to the public on the Internet.

[1642] "Users" are ordinary voters and users who access the website and view the report cards of their representatives.

[1643] This invention relates to a system for evaluating the performance of assembly members and publishing the evaluation results as a “asset member report card.” A specific embodiment of this system will be described below.

[1644] 1. Data collection and storage

[1645] The server collects publicly available minutes and other training data, including official records of parliaments and meetings. The collected data is stored in a MySQL database. Specifically, the evaluation results are inserted into the database using the INSERT statement.

[1646] 2. Data processing

[1647] The server retrieves the evaluation results from the MySQL database. Specifically, it extracts the required data using a SELECT statement. The retrieved data is converted into a data frame using the Python Pandas library and formatted as needed.

[1648] 3. Report Card Generation

[1649] The server uses the Jinja2 template engine to generate an HTML report card based on the processed data. Specifically, it generates the HTML by embedding the data in a template file. The generated HTML file is saved in the public directory of the web server.

[1650] 4. Publication of report cards

[1651] The server uses the Flask framework to build a web server. An endpoint for displaying report cards is set up and made accessible to users. A user accesses the website and views the report card of a specific legislator. Specifically, by entering the legislator's name in the search bar and clicking the search button, the report card of the corresponding legislator is displayed.

[1652] Specific examples

[1653] For example, consider a case where a user accesses a website and wants to view the report card of a particular politician. The user enters the politician's name in the website's search bar and clicks the search button. The server retrieves the evaluation results of the relevant politician from the database and generates and displays the report card in HTML format.

[1654] Prompt Sentence Examples

[1655] "Write a program that generates an HTML report card based on the evaluation results of a specific council member and publishes it on a website. The evaluation results are stored in a MySQL database, and you will use Python's Pandas library and the Jinja2 template engine to process the data and generate the report card. You will then publish the report card on the website using the Flask framework."

[1656] The above is a specific embodiment for carrying out the present invention.

[1657] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1658] Step 1:

[1659] The server collects publicly available minutes and other data as training data. Specifically, it obtains official records of parliaments and meetings through APIs and manual input forms. Input data includes what council members said, attendance rates, and activities. This data is stored in a MySQL database. Specifically, it inserts the evaluation results into the database using INSERT statements. The output is the evaluation results stored in the database.

[1660] Step 2:

[1661] The server retrieves the evaluation results from the MySQL database. Specifically, it uses a SELECT statement to extract the necessary data. The input is the evaluation results stored in the database. The retrieved data is converted into a data frame using the Python Pandas library. The converted data frame is then formatted as needed. For example, the evaluation scores are sorted in ascending order. The output is a formatted data frame.

[1662] Step 3:

[1663] The server uses the Jinja2 template engine to generate an HTML report card based on the processed data. The input is a formatted data frame. Specifically, HTML is generated by embedding data in a template file. The generated HTML file is saved in the public directory of the web server. The output is the generated HTML report card.

[1664] Step 4:

[1665] The server uses the Flask framework to build a web server. An endpoint for displaying report cards is set up. The input is the generated report card in HTML format. A user accesses the website and wants to view the report card of a specific legislator. Specifically, they enter the legislator's name in the search bar and click the search button. The server receives the user's request and displays the report card of the relevant legislator. The output is the report card displayed in the user's browser.

[1666] (Application example 2)

[1667] Next, a description will be given of Application Example 2 of Form Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1668] In conventional evaluation systems, it was difficult for users to easily refer to evaluation results based on the target's activities, comments, attendance rate, etc., and use them as a reference for decision-making. In addition, there was a lack of a way to save the evaluation results in a database and publish them in HTML format, making it difficult for users to visually check the evaluation results and post reviews.

[1669] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes a means for the artificial intelligence to learn using published minutes and the like as training data, a means for the learned artificial intelligence to quantitatively and qualitatively evaluate the target's performance, a means for creating and publishing the evaluation results as a "report card," a means for saving the evaluation results in a database and generating the report card in HTML format, and a means for users to refer to the evaluation results and post reviews. This allows users to visually check the evaluation results and use them as a reference for decision-making.

[1670] "Published minutes, etc." refers to minutes and related documents that are made publicly accessible.

[1671] "Artificial intelligence" refers to systems that mimic human intelligence through machine learning and data analysis.

[1672] "Means of learning" refers to the process by which artificial intelligence analyzes data based on publicly available minutes and other documents to find patterns and rules.

[1673] "Quantitative and qualitative evaluation methods" refers to the process of evaluating the functioning of an object using numerical and qualitative data.

[1674] A "report card" refers to a report-style document summarizing evaluation results.

[1675] "Disclosure means" refers to the process of making evaluation results publicly accessible.

[1676] "Means for storing in a database" refers to the process of storing the evaluation results in a structured format in a database.

[1677] "Means for generating an HTML report card" refers to the process of generating the evaluation results as an HTML document.

[1678] "Means for users to view rating results and post reviews" refers to an interface that allows users to view rating results and add their own opinions and ratings.

[1679] To implement the present invention, the following system configuration is required.

[1680] System Configuration

[1681] 1. Server

[1682] Hardware: A server with a powerful processor, sufficient memory, and storage.

[1683] Software: Python, Flask, SQLite, generative AI models.

[1684] 2. Terminal

[1685] Hardware: Smartphone.

[1686] Software: Web browser or dedicated application.

[1687] 3. Users

[1688] Role: View ratings and post reviews.

[1689] Program processing explanation

[1690] server

[1691] The server provides a means for the AI ​​to learn from publicly available meeting minutes and other data as training data. Specifically, the generative AI model is used to analyze the meeting minutes data and quantitatively and qualitatively evaluate the target's performance. The evaluation results are stored in a database and generated as a report card in HTML format. This allows users to visually check the evaluation results and use them as a reference for decision-making.

[1692] Terminal

[1693] The terminal provides an interface that allows users to view the evaluation results and post reviews. Using their smartphones, users can view the HTML report card provided by the server and add their own opinions and evaluations.

[1694] Specific examples

[1695] For example, when a user opens the app on their smartphone, they are presented with a list of nearby stores' ratings. Each store's rating includes reviews and star ratings posted by other users. Users can post new reviews, which are then sent to the server and stored in a database.

[1696] Prompt Sentence Examples

[1697] "Create a smartphone app that displays the ratings of nearby businesses and allows users to post reviews. The ratings will be stored in a database and displayed in HTML format."

[1698] The above is a specific embodiment for carrying out the present invention.

[1699] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1700] Step 1:

[1701] The server collects published minutes and stores them in a database. The input is the published minutes and other data, and the output is the raw data stored in the database. Specifically, the minutes data is obtained using web scraping and APIs and stored in an SQLite database.

[1702] Step 2:

[1703] The server uses a generative AI model to analyze the minutes data stored in the database and quantitatively and qualitatively evaluate the subject's performance. The input is the minutes data stored in the database, and the output is the evaluation results. Specifically, it uses natural language processing technology to analyze the minutes data and evaluate each subject's activities, comments, attendance rate, etc.

[1704] Step 3:

[1705] The server saves the evaluation results in a database and generates a report card in HTML format. The input is the evaluation results, and the output is an HTML report card. Specifically, it generates an HTML template based on the evaluation results and saves it in the database.

[1706] Step 4:

[1707] The server publishes the report card in HTML format through a web server so that users can view the evaluation results. The input is the report card in HTML format, and the output is the report card displayed in the user's browser. Specifically, a web server is built using Flask so that users can access it.

[1708] Step 5:

[1709] Users access the server using a smartphone browser or a dedicated app to view the evaluation results. The input is the user's access request, and the output is a display of the evaluation results. Specifically, the user opens a browser and accesses the server's URL, and the report card is displayed.

[1710] Step 6:

[1711] Users post reviews based on the evaluation results. The input is the user's review content, and the output is the review sent to the server. Specifically, the user enters the review in the form and presses the submit button, which sends the review to the server.

[1712] Step 7:

[1713] The server stores the reviews sent by users in a database and makes them available for other users to refer to. The input is the user's review content, and the output is the review saved in the database. Specifically, the review content is stored in the database and reflected in an HTML report card.

[1714] The above are the specific processing steps for carrying out the present invention.

[1715] Example 3

[1716] Next, a third embodiment of the third embodiment will be described. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1717] In conventional elections, it is difficult for voters to fully understand the activities and evaluations of their representatives, resulting in a lack of information to decide who to vote for. In addition, there is a lack of means to effectively utilize publicly available parliamentary records and quantitatively and qualitatively evaluate the activities of representatives, making it difficult for voters to decide who to vote for based on reliable information.

[1718] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[1719] In this invention, the server includes means for the artificial intelligence to learn using publicly available parliamentary records and the like as training data, means for the trained artificial intelligence to quantitatively and qualitatively evaluate the activities of legislators, means for creating and publishing the evaluation results as "legislator report cards," means for displaying the evaluation results on a website, and means for users to refer to the evaluation results on the website and decide who to vote for. This enables voters to evaluate the activities of legislators based on reliable information and decide who to vote for.

[1720] "Public parliamentary records" are information such as minutes, attendance records, and statements made that are officially recorded in parliament and made available to the public.

[1721] "Artificial intelligence" is a system that uses machine learning and data analysis techniques to learn patterns from data and make predictions and classifications.

[1722] "The activities of a member of parliament" are the actions and duties of a member of parliament, such as attending parliament, speaking, proposing bills, and voting.

[1723] "Quantitative and qualitative evaluation" means conducting both quantitative evaluation using numerical data and qualitative evaluation that evaluates content and quality.

[1724] A "legislator report card" is a report summarizing the activities and evaluation results of a member of parliament, and is a resource that helps voters understand the performance of their members.

[1725] A "website" is an online page published on the Internet that users can access and view information.

[1726] A "user" is a voter who uses the system to obtain information and decide how to vote.

[1727] "Training data" is a dataset consisting of known input and output pairs that is used to train an artificial intelligence.

[1728] "Learning" is the process by which artificial intelligence finds patterns and rules based on training data, improving the accuracy of predictions and classifications.

[1729] The "evaluation results" are the results of the AI ​​analyzing the activities of lawmakers and evaluating them quantitatively and qualitatively.

[1730] This invention is a system that provides evaluation results that voters can use as a reference when deciding who to vote for. Specifically, a server collects publicly available parliamentary records, evaluates the activities of legislators using artificial intelligence, and displays the evaluation results on a website. Users can refer to the evaluation results on the website and decide how to vote appropriately.

[1731] Hardware and software used

[1732] 1. Server

[1733] The server collects data, processes the data, generates evaluation results, and displays them on the website.

[1734] Software used includes Python, BeautifulSoup, Pandas, Scikit-learn, Flask or Django.

[1735] 2. Terminal

[1736] The terminal is used by the user to access the website and view the evaluation results.

[1737] The hardware used includes computers, smartphones, tablets, etc.

[1738] Software used includes a web browser (e.g., Google Chrome, Mozilla Firefox, Safari).

[1739] 3. Users

[1740] Users access the website as voters and decide who to vote for based on the evaluation results.

[1741] Program processing

[1742] 1. Data Collection

[1743] The server collects data such as each member's attendance rate and speech content from the parliament's official website and publicly available minutes.

[1744] Example: The server uses Python's BeautifulSoup library to scrape HTML data from the official parliamentary website, for example, parsing the page listing the attendance rates of members of parliament and extracting the required data.

[1745] 2. Data processing

[1746] The server organizes the collected data and creates a report card for each member of parliament, including attendance rate, evaluation of speech content, number of bills proposed, and more.

[1747] Example: The server uses the Pandas library to convert the collected data into a data frame. Then, it extracts the necessary information for each member of parliament and formats it into a report card format. For example, it extracts the attendance column and stores it in a new data frame along with each member's name.

[1748] 3. Generating evaluation results

[1749] The server generates evaluation results for each member of parliament, which are scored based on their attendance rate and the content of their comments.

[1750] Example: The server uses the Scikit-learn library to build a model for scoring the assessment results, for example a linear regression model that assigns higher scores to students with higher attendance rates.

[1751] 4. Display on the Website

[1752] The server displays the generated evaluation results on a website, and users can access the website to view the report cards of each member of parliament.

[1753] Example: The server uses a web framework such as Flask or Django to create a web page that displays the evaluation results. For example, it generates an HTML page that displays the names of each legislator and their evaluation scores in a table.

[1754] 5. User voting decision

[1755] Users can refer to the report cards of each member of parliament on the website and decide who to vote for based on the evaluation results.

[1756] Example: A user accesses a website using a browser, checks the displayed ratings, and, for example, selects the legislator with the highest rating score and decides to vote for that legislator.

[1757] Prompt Sentence Examples

[1758] Examples of prompts to be input to a generative AI model include:

[1759] "In order to decide which representative to vote for in the next election, I would like to refer to a report card based on the attendance rate and comments of each representative. I would like to check the evaluation results on the website and choose the representative with the highest evaluation."

[1760] By using this prompt sentence, the generative AI model can understand what information the user is looking for and provide an appropriate evaluation result. The flow of the identification process in Example 3 will be described with reference to Figure 15.

[1761] Step 1:

[1762] Data collection

[1763] The server collects data such as each member's attendance rate and speech content from the parliament's official website and publicly available minutes.

[1764] Input: URL of the official parliament website or publicly available minutes.

[1765] What it does: The server uses Python's BeautifulSoup library to scrape HTML data from the official parliamentary website. For example, it parses the page listing the attendance rates of members of parliament and extracts the necessary data.

[1766] Output: Raw data such as attendance rate and speech content of each member of parliament.

[1767] Step 2:

[1768] Data Processing

[1769] The server organizes the collected data and creates a report card for each member of parliament, including attendance rate, evaluation of speech content, number of bills proposed, and more.

[1770] Input: The raw data collected in step 1.

[1771] What it does: The server uses the Pandas library to convert the collected data into a data frame. Then, it extracts the necessary information for each member of parliament and formats it into a report card format. For example, it extracts the attendance column and stores it in a new data frame along with each member's name.

[1772] Output: Report card data for each member of parliament.

[1773] Step 3:

[1774] Generating evaluation results

[1775] The server generates evaluation results for each member of parliament, which are scored based on their attendance rate and the content of their comments.

[1776] Input: Report card data created in step 2.

[1777] What it does: The server uses the Scikit-learn library to build a model for scoring the assessment results, for example, a linear regression model that assigns higher scores to students with higher attendance rates.

[1778] Output: Evaluation scores for each legislator.

[1779] Step 4:

[1780] Display on the website

[1781] The server displays the generated evaluation results on a website, and users can access the website to view the report cards of each member of parliament.

[1782] Input: The rating scores generated in step 3.

[1783] What it does: The server uses a web framework such as Flask or Django to create a web page that displays the evaluation results. For example, it generates an HTML page that displays the names of each legislator and their evaluation scores in a table.

[1784] Output: A web page that displays the evaluation results.

[1785] Step 5:

[1786] User voting decision

[1787] Users can refer to the report cards of each member of parliament on the website and decide who to vote for based on the evaluation results.

[1788] Input: The assessment results displayed on the website.

[1789] Specific operation: A user accesses a website using a browser, checks the displayed evaluation results, and, for example, selects the legislator with the highest evaluation score and decides to vote for that legislator.

[1790] Output: User's vote decision.

[1791] (Application example 3)

[1792] Next, a description will be given of Application Example 3 of Form Example 3. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1793] In conventional lawmaker evaluation systems, evaluation results are simply published on a website, and users lack sufficient interactivity and convenience when deciding who to vote for based on that information. Furthermore, there is a lack of means to provide information in physical stores, and there is no system in place to allow users to check evaluation results in real time and receive recommendations on the optimal candidate to vote for.

[1794] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[1795] In this invention, the server includes means for AI to learn using published Diet minutes and the like as training data, means for the trained AI to quantitatively and qualitatively evaluate the performance of Diet members, means for creating and publishing the evaluation results as "Diet member report cards," means for linking with interactive displays in physical stores and displaying information on smartphones using QR codes, means for recommending optimal voting destinations to users based on the evaluation results, and means for updating the evaluation results and recommendation information in real time. This enables users to interactively check the evaluation results in physical stores and receive optimal voting destination recommendations in real time.

[1796] "Published Diet minutes" are documents that record the discussions and statements made in the Diet and are available to the public.

[1797] "Teacher data" is a dataset used for learning by a machine learning model, and is data that has been assigned a correct answer label.

[1798] "AI" is an abbreviation for artificial intelligence, a technology that allows computers to imitate human intelligence to learn and reason.

[1799] "The work of a member of parliament" is a general term for the activities, statements, attendance, etc. that members of parliament undertake in the Diet and other official duties.

[1800] "Quantitative and qualitative evaluation" means evaluation using both numerical data (quantitative) and qualitative data (qualitative).

[1801] A "Member of Parliament Report Card" is a report summarizing the evaluation results of a member of parliament's activities, comments, attendance rate, etc.

[1802] An "interactive display in a physical store" is a display device that can be directly operated by a user and is installed in a physical store.

[1803] A "QR code" is a type of two-dimensional barcode that can be read by devices such as smartphones.

[1804] A "smartphone" is a mobile phone that can connect to the Internet and use applications.

[1805] "Recommending the best voting option" means suggesting the most suitable voting option for the user.

[1806] "Real-time updates" means that data and information are updated immediately and always kept up to date.

[1807] The system for implementing this invention is configured as follows: First, the server has a means for the AI ​​to learn from publicly available Diet minutes and other training data. This AI is used to quantitatively and qualitatively evaluate the performance of Diet members. The evaluation results are compiled as "Diet member report cards" and made public.

[1808] The second step involves linking the system to interactive displays in physical stores and displaying information on smartphones using QR codes. Users can check the evaluation results of each legislator by scanning the QR code displayed on the display in the store with their smartphone.

[1809] Furthermore, the server has a means to recommend the best vot...

Claims

1. A means for artificial intelligence to learn from the published minutes as training data, and A means for the learned artificial intelligence to quantitatively and qualitatively evaluate the work of lawmakers; A means for creating a "report card" of the evaluation results of said legislators; means for outputting the evaluation results in a format viewable by a user; A system including:

2. The system according to claim 1, wherein the evaluation results are calculated based on the activities, statements, and attendance rate of the members of parliament.

3. The system according to claim 1 , wherein the evaluation results are made public in a form that can be referenced by users, and serve as reference information for users in deciding who to vote for.

Images