System

Abstract

A system is provided.SOLUTION: A system, comprising: means for obtaining a plurality of submitted proposals; means for converting the proposals into numerical vectors; means for calculating a similarity of each proposal in a vector space; and means for extracting a proposal whose similarity is less than or equal to a certain threshold based on the calculated similarities.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] In innovation contests and idea pitch events, there is a need to efficiently find innovative and unique proposals from among a large number of proposals, but conventional methods have the problem of being extremely time-consuming and labor-intensive.The objective of this invention is to solve this problem and provide a system that can efficiently extract unique proposals. [Means for solving the problem]

[0005] The present invention solves the above-mentioned problems by providing a system including a means for acquiring multiple submitted proposals, a means for converting the proposals into numerical vectors, a means for calculating the similarity of each proposal in a vector space, and a means for extracting proposals whose similarity is below a certain threshold based on the calculated similarity. By using cosine similarity to calculate the similarity and TF-IDF or Word2Vec to convert the proposals into numerical vectors, the innovativeness of the proposals can be efficiently and automatically evaluated and extracted.

[0006] A "proposal" is a new idea or plan submitted by a participant in an innovation contest or idea pitch event.

[0007] A "numeric vector" is a multidimensional data structure for representing text data as numbers, where each dimension represents a specific feature and the whole describes the content of the suggestion.

[0008] A "vector space" is a multidimensional mathematical space in which numerical vectors are arranged, allowing distances and angles between vectors to be calculated.

[0009] "Similarity" is an index that shows the relationship between two numerical vectors. It is calculated using cosine similarity, etc.

[0010] "Cosine similarity" is a method for measuring similarity using the cosine value of the angle between two vectors. The closer this value is to 1, the higher the similarity.

[0011] "TF-IDF" stands for "Term Frequency-Inverse Document Frequency" and is a technology for extracting and evaluating important words in text data. It is calculated based on the frequency and inverse document frequency of each word.

[0012] "Word2Vec" is a technology for converting words into numerical vectors, and is a method for numerically expressing the meaning of words.

[0013] A "threshold" is a reference value set for evaluating the similarity score. Based on this value, the similarity is judged to be high or low.

[0014] "Extraction" refers to the process of extracting proposals that meet specific conditions from among a large number of proposals. [Brief explanation of the drawings]

[0015] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13]FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0016] 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.

[0017] First, the terms used in the following description will be explained.

[0018] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0019] 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.

[0020] 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.

[0021] 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.

[0022] 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."

[0023] [First embodiment]

[0024] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0025] 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.

[0026] 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).

[0027] 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.

[0028] 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.

[0029] 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.

[0030] 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.

[0031] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0032] 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.

[0033] 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.

[0034] 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.

[0035] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0036] This invention relates to a system that efficiently analyzes multiple proposals submitted in innovation contests and idea pitch events and automatically extracts innovative proposals. This system includes the following processing steps, and can consistently perform proposal submission, processing, and evaluation.

[0037] Program processing procedure

[0038] 1. Idea collection

[0039] Users submit ideas to innovation contests and idea pitch events via web forms or dedicated applications.

[0040] 2. Save your ideas in a database

[0041] The server receives the submitted ideas and stores them in a database. The stored data includes the idea text, submitter information, and submission date and time.

[0042] 3. Preprocessing ideas using natural language processing

[0043] The server retrieves unprocessed ideas from the database and performs the following natural language processing on the retrieved ideas.

[0044] Stop word removal: remove common meaningless words (e.g. "no", "wa", "ni", etc.).

[0045] Stemming: Extracting the base form of a word. (Example: "hashiru" → "hashiru")

[0046] Keyword extraction: Extract important words and phrases.

[0047] 4. Converting ideas into numerical vectors

[0048] The server converts the preprocessed idea text data into a numerical vector using the following techniques:

[0049] TF-IDF (Term Frequency-Inverse Document Frequency): Evaluates and quantifies the importance of each word.

[0050] Word2Vec: Generates numeric vectors based on word meanings.

[0051] 5. Similarity calculation in vector space

[0052] The server calculates the cosine similarity between the numeric vectors of all ideas, which allows it to calculate a similarity score between proposals and identify less relevant proposals.

[0053] 6. Extracting unique ideas

[0054] The server extracts ideas with similarity scores below a certain threshold. These are considered unique ideas that are significantly different from other proposals and are stored in a database as "unique ideas."

[0055] 7. Notices and Confirmations

[0056] The server notifies the judges and users, prompting them to review and evaluate the extracted unique ideas.

[0057] Specific examples

[0058] Example: Idea submission and processing

[0059] 1. A user submits two ideas using a web form, for example, "a new energy-efficient battery design" and "furniture made from recyclable materials."

[0060] 2. The server stores these ideas in a database and marks them as pending.

[0061] 3. The server performs natural language processing to remove stop words, stem, and extract key keywords from each idea.

[0062] Idea A Keywords: ["energy", "efficiency", "battery", "design"]

[0063] Idea B Keywords: ["recyclable", "material", "furniture"]

[0064] 4. The server generates a numeric vector based on the keyword.

[0065] Idea A's vector: [0.5, 0.7, 0.6, 0.8]

[0066] Idea B vector: [0.1, 0.3, 0.9, 0.4]

[0067] 5. The server calculates the cosine similarity and finds that the similarity score between these ideas is 0.2. Based on this low score, Idea A and Idea B are extracted as unique ideas.

[0068] 6. The server notifies the judges and prompts them to review and evaluate the extracted ideas.

[0069] This system makes it possible to efficiently extract innovative and unique ideas from a large number of proposals.

[0070] The processing flow will be explained below.

[0071] Step 1:

[0072] A user enters their idea in a web form or a dedicated application and clicks the submit button. The idea is entered in text format.

[0073] Step 2:

[0074] The terminal transmits the idea input by the user to the server, and the transmitted data includes the idea text, the user's identification information, the submission date and time, etc.

[0075] Step 3:

[0076] The server stores the received idea data in the database, and the stored data is marked with an unprocessed flag for the idea.

[0077] Step 4:

[0078] The server retrieves ideas with unprocessed flags from the database, and the retrieved ideas are subjected to natural language processing.

[0079] Step 5:

[0080] The server removes stop words from the idea text, which in the case of Japanese is a process to remove common words such as "no", "wa", and "ni".

[0081] Step 6:

[0082] The server performs stemming and converts words to their base form, for example, "run" and "ran" are converted to "run".

[0083] Step 7:

[0084] The server performs keyword extraction and lists important words and phrases, which are later converted into numerical vectors.

[0085] Step 8:

[0086] The server takes the preprocessed text and converts it into a numerical vector, using techniques such as TF-IDF or Word2Vec.

[0087] Step 9:

[0088] The server calculates the cosine similarity between the generated numeric vectors for all ideas. This similarity score evaluates the relative relevance of the proposals.

[0089] Step 10:

[0090] The server evaluates the similarity scores and extracts ideas with scores below a set threshold, which is used to identify ideas that are unique and different from other proposals.

[0091] Step 11:

[0092] The server stores the extracted unique ideas in a database and sets a special flag so that they are given priority access in subsequent evaluation processes.

[0093] Step 12:

[0094] The server sends notifications to users and judges, including a summary of the unique ideas extracted and a request for confirmation and evaluation.

[0095] Example 1

[0096] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0097] Conventional proposal extraction systems have had difficulty efficiently extracting innovative and unique proposals from a large number of proposals. Furthermore, the accuracy of proposal preprocessing and similarity calculations was insufficient, making it impossible to guarantee the quality of the extracted proposals. As a result, proposal evaluation took a long time, increasing the burden on judges and users.

[0098] 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.

[0099] In this invention, the server includes means for acquiring a plurality of submitted proposals, means for storing the acquired proposals in a database, means for acquiring unprocessed proposals from the database, means for preprocessing text data of the proposals using natural language processing, means for converting the preprocessed proposals into numerical vectors, means for calculating the similarity of each proposal in a vector space, means for extracting proposals whose similarity is equal to or less than a certain threshold based on the calculated similarity, and means for notifying the extracted proposals. This makes it possible to efficiently extract innovative and unique proposals from a large number of proposals.

[0100] "Proposal" means an idea or outline submitted by a User in an innovation contest or idea pitch event.

[0101] "Database" refers to a collection of information for storing and managing submitted proposals and related information (such as submitter information, submission date and time, etc.).

[0102] "Natural language processing" is a general term for technology that analyzes text data, understands its meaning, and extracts necessary information.

[0103] A "numeric vector" is a mathematical representation of text data or proposal content in a format that allows for calculations within a vector space.

[0104] "Cosine similarity" is a calculation method for evaluating the similarity between two numerical vectors, and uses the angle between the vectors.

[0105] "TF-IDF" stands for "Term Frequency-Inverse Document Frequency" and is a method for calculating and quantifying the importance of each word.

[0106] "Word2Vec" is a technology for representing the meaning of words as numerical vectors, and is a model that trains words that are semantically similar to each other so that they have similar vectors.

[0107] "Notification" refers to the action of sending information so that judges and users are aware of the proposal.

[0108] The present invention relates to a system that efficiently analyzes multiple proposals submitted in innovation contests and idea pitch events and automatically extracts innovative proposals. This system operates using the following hardware and software.

[0109] Hardware and software used

[0110] Server: A computer system for database management, natural language processing, numeric vector transformation, similarity calculation, and notification sending.

[0111] Database: An information management system for storing proposal data, user information, and similarity scores.

[0112] Natural language processing libraries: For example, use Python's nlp library to analyze text data.

[0113] Numerical vector transformation libraries: For example, use the scikit-learn library to perform TF-IDF and Word2Vec transformations.

[0114] Similarity calculation library: For example, use the scipy library to calculate cosine similarity.

[0115] Notification Sending API: API for sending emails and push notifications.

[0116] Example of a system

[0117] Example: Idea submission and processing

[0118] 1. A user submits an idea for a "new energy-efficient battery design" using a web form.

[0119] 2. The server receives the idea and stores it in the database in the format "Idea: New energy-efficient battery design, Submitted by: Username, Submitted date: 2023-10-01".

[0120] 3. The server retrieves raw ideas from the database and performs natural language processing, such as removing "no," "wa," and "ni" from the idea text, stemming "hashiru" to "so," and extracting keywords such as "energy," "efficiency," "battery," and "design."

[0121] 4. The server calculates the TF-IDF based on these keywords and generates a numerical vector, for example, "energy" is 0.7, "efficiency" is 0.8, "battery" is 0.9, and "design" is 1.0.

[0122] 5. The server calculates the cosine similarity between all ideas and obtains a similarity score (e.g., 0.2).

[0123] 6. The server extracts ideas with a similarity score below a threshold (e.g., 0.3) as unique ideas and stores them in the database. For example, it flags them as "unique ideas" using an SQL UPDATE statement.

[0124] 7. The server sends a notification email to the judges, prompting them to confirm and evaluate the unique ideas.

[0125] Example prompts for generative AI models

[0126] Here are some example prompts you can enter into your generative AI model:

[0127] "Please extract important keywords from the following text: 'New energy-efficient battery design'"

[0128] The present invention makes it possible to efficiently extract innovative and unique proposals from a large number of proposals, thereby improving the efficiency of the review and the quality of the evaluation.

[0129] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0130] Step 1:

[0131] A user submits an idea. Using a web form, they enter the title and body of the idea and click the "Submit" button. The input is the title and body of the idea, and the output is the idea sent to the server.

[0132] Step 2:

[0133] The server receives the submitted idea. This temporarily stores the submitted idea data (title, text, submitter information, submission date and time). The input is the idea data sent by the user, and the output is the data stored in the server's temporary memory.

[0134] Step 3:

[0135] The server saves the received idea data in the database. The server establishes a database connection and uses the SQL INSERT statement to store the idea data in persistent storage. As input, there is idea data stored in temporary memory, and as output, the idea data is saved in the database.

[0136] Step 4:

[0137] The server retrieves unprocessed ideas from the database. It uses a SQL SELECT statement to extract idea data that is marked as unprocessed. As input, it has idea data that is flagged as unprocessed, and as output, it extracts that data to the server.

[0138] Step 5:

[0139] The server performs natural language processing (NLP) to preprocess the idea data. It uses Python's NLP library to remove stop words, perform stemming, and extract keywords. The extracted idea data is input, and preprocessed idea data is generated as output. Specifically, it removes words such as "no," "wa," and "ni" from the main body of the idea, converts "hashiru" to "so," and extracts keywords such as "energy," "efficiency," "battery," and "design."

[0140] Step 6:

[0141] The server converts the preprocessed idea data into a numerical vector. Using Python's scikit-learn library, the conversion is performed using TF-IDF and Word2Vec. The preprocessed idea data is input, and a numerical vector is generated as output. Specifically, the importance of each keyword is evaluated and quantified.

[0142] Step 7:

[0143] The server calculates the cosine similarity between all idea vectors. The similarity calculation is performed using the scipy library. The input is a set of numerical vectors, and the output is a similarity score between each vector. Specifically, the angle between the vectors is calculated and the result is scored.

[0144] Step 8:

[0145] The server extracts unique ideas based on the similarity score. The server stores ideas with a similarity score below a certain threshold as "unique ideas" in a database. The input is the similarity score, and the output is data flagged as a unique idea.

[0146] Step 9:

[0147] The server sends notifications to judges and users. The server uses an email sending API to notify unique ideas that require evaluation. The input is the data of the unique idea, and the output is a notification email. Specifically, the body of the notification email contains the title of the idea and an evaluation link.

[0148] (Application example 1)

[0149] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0150] Proposing suggestions to improve work efficiency on-site requires a lot of time and effort. In particular, the process of analyzing information collected from the factory floor, extracting useful suggestions, and quickly communicating the results to workers is complex, which can delay responses when immediate action is required. In addition, traditional methods often include many overlapping or similar suggestions, which can result in useful suggestions being overlooked.

[0151] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0152] In this invention, the server includes a means for acquiring a plurality of submitted proposals, a means for converting the proposals into a numerical vector, a means for calculating similarity, a means for extracting proposals whose similarity is equal to or less than a certain threshold, and a means for using a smart device to present the extracted proposals. This makes it possible to quickly identify useful proposals for supporting improved work efficiency at the factory site and quickly notify workers.

[0153] The "means for acquiring a plurality of submitted proposals" refers to a means by which workers or engineers on the factory floor use smart devices to input proposals and transmit them to a server.

[0154] The "means for converting into a numerical vector" is a means for converting the proposed text data into a numerical vector using a specific algorithm (TF-IDF or Word2Vec).

[0155] The "means for calculating similarity" is a means for performing calculations to evaluate the similarity between proposals using the converted numerical vectors.

[0156] The "means for extracting proposals whose similarity is equal to or less than a certain threshold" is a means for identifying and extracting unique proposals that are significantly different from other proposals, based on the similarity scores between proposals.

[0157] The "means of using a smart device to present extracted suggestions" refers to a means of notifying and displaying the extracted unique suggestions to workers or engineers in real time through a smart device such as smart glasses.

[0158] The present invention is a proposed analysis system aimed at improving work efficiency at a factory site. A specific embodiment of this system is described below.

[0159] First, workers or engineers use smart glasses to voice-input improvements or new ideas they discover on-site. The voice-input content is converted into text data by the voice recognition software in the smart glasses. This text data is then sent to a server via a network. This process uses smart glasses (e.g., Google Glass) and their built-in voice recognition software (e.g., Google Speech-to-Text API).

[0160] The server stores the received suggestions in a database and then performs the following natural language processing: The server removes common meaningless words (stop words) from the suggested text and converts words to their root forms (stemming). This process is performed using the NLTK library. After that, the text data is converted into a numerical vector using the TF-IDF (Term Frequency-Inverse Document Frequency) or Word2Vec algorithm to extract important keywords. This is performed using the Scikit-learn library.

[0161] Next, the server calculates the cosine similarity between the vectors and calculates a similarity score for each proposal. It identifies proposals with a similarity score below a certain threshold (unique proposals) and extracts them. The Scikit-learn library is also used to calculate the similarity.

[0162] The extracted unique suggestions are sent to workers and engineers using smart glasses. The notification includes an overview of the suggestion and its evaluation, allowing workers to immediately check and evaluate the content of the suggestion. This enables quick responses on-site and is expected to lead to efficient work improvements.

[0163] As a specific use case, field worker A makes a suggestion by voice, "Design a new, energy-efficient battery." This suggestion is sent to the server, where it undergoes natural language processing and similarity calculations are performed. Similarly, worker B makes a suggestion, "Furniture made from recyclable materials." The server evaluates the similarity between these suggestions and identifies both as unique suggestions. This allows workers to instantly receive useful suggestions and incorporate them into their improvement activities.

[0164] An example of an input prompt for the generative AI model is as follows:

[0165] Next, enter the idea you want to collect. For example, enter a brief description of your idea, such as "new energy-efficient battery designs" or "furniture made from recyclable materials."

[0166] By using the above-described means, the present invention improves the efficiency of a series of processes from collection of proposals to analysis and notification at the factory site, thereby contributing to improved work efficiency.

[0167] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0168] Step 1:

[0169] The user uses the smart glasses to speak suggestions, which are converted into text by voice recognition software within the glasses, and then sent to the server.

[0170] Input: User's voice suggestions

[0171] Output: Proposals converted to text data

[0172] How it works: The smart glasses record the voice suggestions and convert them into text using voice recognition software, which then sends the text data over the network to a server.

[0173] Step 2:

[0174] The server stores the received proposal in a database. Data including details of the proposal (text, submitter information, submission date and time, etc.) is recorded in the database.

[0175] Input: Text data

[0176] Output: Proposal data stored in a database

[0177] What happens: The server receives the text data and stores it as a new record in the database. All necessary metadata is also saved at this stage.

[0178] Step 3:

[0179] The server retrieves the raw stored suggestions and performs natural language processing on them, specifically stop word removal, stemming, and keyword extraction.

[0180] Input: Proposal data retrieved from the database

[0181] Output: Preprocessed text data

[0182] Specific operation: The server uses the NLTK library to remove stop words and perform stemming, and performs keyword extraction using algorithms such as TF-IDF and Word2Vec.

[0183] Step 4:

[0184] The server converts the preprocessed text data of the proposals into a numerical vector using TF-IDF or Word2Vec.

[0185] Input: Preprocessed text data

[0186] Output: Numeric vector

[0187] Specific operation: The server converts text data into a numerical vector using TF-IDF or Word2Vec from the Scikit-learn library.

[0188] Step 5:

[0189] The server calculates the cosine similarity between all numeric vectors, which then produces a similarity score between the proposals.

[0190] Input: Numeric vector

[0191] Output: Similarity score

[0192] Specific operation: The server uses Scikit-learn's cosine_similarity method to calculate the cosine similarity between numeric vectors and obtain a similarity score.

[0193] Step 6:

[0194] The server extracts proposals with similarity scores below a certain threshold and identifies them as unique proposals.

[0195] Input: Similarity score

[0196] Output: A list of unique suggestions

[0197] Specific operation: The server evaluates the similarity score based on the set threshold and extracts a list of unique proposals that are significantly different from other proposals.

[0198] Step 7:

[0199] The server notifies the user of the extracted unique suggestions through the smart glasses, along with a summary of the suggestions and their ratings.

[0200] Input: A list of unique suggestions

[0201] Output: Notification of the suggestion displayed on the smart glasses

[0202] Specific operation: The server composes the extracted unique ideas into a notification message and sends it to the smart glasses via the network. The smart glasses then display the content of the suggestion to the user.

[0203] Through the above processing steps, the present invention makes it possible to quickly identify useful suggestions for improving work efficiency on the factory floor and notify workers of them.

[0204] 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.

[0205] The present invention achieves even more accurate evaluation by combining a system that analyzes multiple proposals submitted to innovation contests and idea pitch events and automatically extracts innovative proposals with an emotion engine that recognizes the user's emotions.

[0206] Program processing procedure

[0207] 1. Idea collection

[0208] Users submit ideas to innovation contests and idea pitch events via web forms or dedicated applications.

[0209] 2. Save your ideas in a database

[0210] The server receives the submitted ideas and stores them in a database. The stored data includes the idea text, submitter information, and submission date and time.

[0211] 3. Preprocessing ideas using natural language processing

[0212] The server retrieves unprocessed ideas from the database and performs the following natural language processing on the retrieved ideas.

[0213] Stop word removal: remove common meaningless words (e.g. "no", "wa", "ni", etc.).

[0214] Stemming: Extracting the base form of a word. (Example: "hashiru" → "hashiru")

[0215] Keyword extraction: Extract important words and phrases.

[0216] 4. Converting ideas into numerical vectors

[0217] The server converts the preprocessed idea text data into a numerical vector using the following techniques:

[0218] TF-IDF (Term Frequency-Inverse Document Frequency): Evaluates and quantifies the importance of each word.

[0219] Word2Vec: Generates numeric vectors based on word meanings.

[0220] 5. Similarity calculation in vector space

[0221] The server calculates the cosine similarity between the numeric vectors of all ideas, which allows it to calculate a similarity score between proposals and identify less relevant proposals.

[0222] 6. Extracting unique ideas

[0223] The server extracts ideas with similarity scores below a certain threshold. These are considered unique ideas that are significantly different from other proposals and are stored in a database as "unique ideas."

[0224] 7. Leveraging Emotional Engines

[0225] The server uses an emotion engine to acquire the user's emotional data, analyzing facial expressions and voice data when the user submits an idea, and quantifying the user's emotional state.

[0226] 8. Reflecting Emotional Data in Evaluation

[0227] The server influences the evaluation of the proposal based on the emotional data it acquires. For example, if the user has very positive emotions toward a proposal, it will increase the innovativeness rating of that proposal.

[0228] 9. Reevaluate unique ideas

[0229] The server performs re-evaluation using the emotion engine. Based on the emotion data, it recalculates the evaluation of unique ideas that have already been extracted and prioritizes them.

[0230] 10. Notices and Confirmations

[0231] The server sends notifications to users and judges, including a summary of the unique ideas extracted and a request for confirmation and evaluation.

[0232] Specific examples

[0233] Example: Idea submission and processing

[0234] 1. A user submits two ideas using a web form, for example, "a new energy-efficient battery design" and "furniture made from recyclable materials."

[0235] 2. The device sends these ideas to the server.

[0236] 3. The server stores these ideas in a database and marks them as pending.

[0237] 4. The server performs natural language processing to remove stop words, stem, and extract key keywords from each idea.

[0238] Idea A Keywords: ["energy", "efficiency", "battery", "design"]

[0239] Idea B Keywords: ["recyclable", "material", "furniture"]

[0240] 5. The server generates a numeric vector based on the keyword.

[0241] Idea A's vector: [0.5, 0.7, 0.6, 0.8]

[0242] Idea B vector: [0.1, 0.3, 0.9, 0.4]

[0243] 6. The server calculates the cosine similarity and finds that the similarity score between these ideas is 0.2. Based on this low score, Idea A and Idea B are extracted as unique ideas.

[0244] 7. The server uses an emotion engine to extract emotional data from facial expressions and voice when users submit ideas. For example, if a user is very excited, the idea will be rated higher.

[0245] 8. The server re-evaluates and prioritizes ideas based on sentiment data.

[0246] 9. The server notifies the judges and prompts them to review and evaluate the extracted ideas.

[0247] This system efficiently extracts innovative and unique ideas from a large number of proposals, and by evaluating them based on user emotional data, it is possible to achieve highly accurate evaluations.

[0248] The processing flow will be explained below.

[0249] Step 1:

[0250] A user enters their idea in a web form or a dedicated application and clicks the submit button. The idea is entered in text format.

[0251] Step 2:

[0252] The terminal transmits the idea input by the user to the server, and the transmitted data includes the idea text, the user's identification information, the submission date and time, etc.

[0253] Step 3:

[0254] The server stores the received idea data in the database, and the stored data is marked with an unprocessed flag for the idea.

[0255] Step 4:

[0256] The server retrieves ideas with unprocessed flags from the database, and the retrieved ideas are subjected to natural language processing.

[0257] Step 5:

[0258] The server removes stop words from the idea text, which in the case of Japanese is a process to remove common words such as "no", "wa", and "ni".

[0259] Step 6:

[0260] The server performs stemming and converts words to their base form, for example, "run" and "ran" are converted to "run".

[0261] Step 7:

[0262] The server performs keyword extraction and lists important words and phrases, which are later converted into numerical vectors.

[0263] Step 8:

[0264] The server takes the preprocessed text and converts it into a numerical vector, using techniques such as TF-IDF or Word2Vec.

[0265] Step 9:

[0266] The server calculates the cosine similarity between the generated numeric vectors for all ideas. This similarity score evaluates the relative relevance of the proposals.

[0267] Step 10:

[0268] The server evaluates the similarity scores and extracts ideas with scores below a set threshold, which is used to identify ideas that are unique and different from other proposals.

[0269] Step 11:

[0270] The server stores the extracted unique ideas in a database and sets a special flag so that they are given priority access in subsequent evaluation processes.

[0271] Step 12:

[0272] The server uses an emotion engine to acquire emotional data from the user's facial expressions and voice when submitting an idea. The emotional data is analyzed in real time and quantified to represent the user's emotional state (e.g., excitement, joy, surprise, etc.).

[0273] Step 13:

[0274] The server reflects the emotional data in the idea evaluation. Positive emotional data positively influences the evaluation score of the idea, while negative emotional data is used as an indicator to reconsider the evaluation.

[0275] Step 14:

[0276] The server uses the results of the emotion data to re-evaluate the unique ideas that have already been extracted. As a result of the re-evaluation, the priorities may change.

[0277] Step 15:

[0278] The server sends a notification to the user and judges. The notification includes a summary of the unique ideas extracted, as well as the results of a re-evaluation that takes into account the emotional data. The judges then make their final confirmation and evaluation based on this information.

[0279] The following is an explanation based on a specific example.

[0280] Example: Submitting and processing an idea

[0281] Step 1 and Step 2:

[0282] Users submit two ideas through a web form: "A new energy-efficient battery design" and "Furniture made from recyclable materials." The device then sends the ideas to the server.

[0283] Step 3 to Step 7:

[0284] The server stores these ideas in a database and performs natural language processing on each idea, removing stop words, stemming, and extracting important keywords.

[0285] For example, create a list of keywords such as "energy efficient battery design" and "recyclable materials furniture."

[0286] Steps 8 to 10:

[0287] The server converts keywords into numerical vectors and calculates the cosine similarity. Ideas with low similarity scores are extracted as unique ideas and stored in a database.

[0288] Step 11 to Step 14:

[0289] The server uses an emotion engine to obtain user emotion data and reflect it in the idea evaluation. For example, if a user is very excited when submitting a "new energy-efficient battery design," the evaluation score of the idea will be increased.

[0290] Re-evaluation is performed based on the emotional data, and the priority of the extracted unique ideas is recalculated.

[0291] Step 15:

[0292] The server notifies the judges and users and provides an overview of the unique ideas extracted and evaluation results that take into account the emotional data. The judges then make their final confirmation and evaluation based on this information.

[0293] This makes it possible to accurately evaluate and extract ideas that incorporate emotional data.

[0294] Example 2

[0295] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0296] Conventional idea evaluation systems have difficulty efficiently extracting innovative and unique proposals from a large number of proposals. Furthermore, they lack user emotional data, resulting in low evaluation accuracy. A system that can solve these issues and provide more accurate and efficient evaluations is needed.

[0297] 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.

[0298] In this invention, the server includes means for acquiring a plurality of submitted proposals, means for converting the proposals into numerical vectors, means for calculating the similarity of each proposal in a vector space, means for extracting proposals whose similarity is equal to or less than a certain threshold based on the calculated similarity, means for acquiring user emotion data using an emotion engine, and means for recalculating proposal evaluations based on the acquired emotion data. This makes it possible to efficiently extract innovative and unique proposals from a large number of proposals and to enable highly accurate evaluations that reflect the user's emotion data.

[0299] "Submitted Proposals" means ideas or proposals submitted by a User to the System through a contest or event.

[0300] "Means for obtaining" refers to the process or component by which the server receives and stores user input data.

[0301] "Means for converting to a numeric vector" refers to a process or algorithm for converting text data into a numeric vector format using natural language processing techniques.

[0302] "Means for calculating the similarity of each proposal in vector space" refers to a process or algorithm for calculating, for example, the cosine similarity between proposals converted into numerical vectors.

[0303] "Means for extracting proposals with similarity below a certain threshold" refers to a process or algorithm for selecting unique proposals that are significantly different from other proposals based on a calculated similarity score.

[0304] "Means for acquiring user emotional data using an emotion engine" refers to a process or component that uses technologies such as facial expression analysis and voice analysis to detect the user's emotional state and acquire it as data.

[0305] "Means for recalculating the evaluation of a proposal based on acquired emotional data" refers to a process or algorithm for correcting or recalculating the evaluation of the innovativeness or uniqueness of a proposal based on the emotional data.

[0306] This invention relates to a system that analyzes multiple proposals submitted to innovation contests and idea pitch events and automatically extracts innovative proposals. This system achieves more accurate evaluation by combining it with an emotion engine that recognizes the user's emotions.

[0307] The system uses the following hardware and software:

[0308] Server: Receives, stores, analyzes, and evaluates proposal data.

[0309] Terminal: Runs a web form or dedicated application that allows users to submit ideas.

[0310] Emotion engine: Facial expression and voice analysis technology to recognize user emotions (e.g., facial expression recognition software, voice recognition software).

[0311] System hardware and software configuration

[0312] 1. A means of retrieving multiple submitted proposals

[0313] Users submit ideas through a web form or a dedicated application, for example, by entering an idea such as "Design a new, energy-efficient battery."

[0314] 2. A means of converting proposals into numerical vectors

[0315] The server uses natural language processing technology to convert the text data of submitted ideas into numerical vectors, specifically using TF-IDF and Word2Vec to quantify the importance and meaning of words.

[0316] 3. A means to calculate the similarity of each proposal in the vector space

[0317] The server calculates the cosine similarity between the proposals converted into numerical vectors, which allows us to evaluate the relevance between the proposals.

[0318] 4. A method for extracting proposals whose similarity is below a certain threshold

[0319] The server extracts unique proposals that are significantly different from other proposals based on the calculated similarity scores, thus identifying innovative and unique proposals.

[0320] 5. Means of acquiring user emotion data using an emotion engine

[0321] The server uses an emotion engine to analyze facial expressions and voice data when a user submits a suggestion and quantify their emotional state. For example, a user who is very excited might be recorded as 80%.

[0322] 6. A method for recalculating the evaluation of proposals based on acquired emotional data

[0323] The server reflects the sentiment data and recalculates the evaluation of the proposal's innovativeness and uniqueness, so that the user's sentiment influences the proposal's final evaluation.

[0324] As a concrete example, consider a case where a user submits two ideas using a web form, say "designing a new energy-efficient battery" and "furniture made from recyclable materials." The device sends these ideas to a server, which stores them in a database and marks them as unprocessed. The server then performs natural language processing (NLP) to remove stop words from each idea, perform stemming, and extract key keywords. The server then generates a numerical vector based on the keywords and calculates the cosine similarity to calculate a similarity score between the proposals.

[0325] A particularly important step is to use an emotion engine to obtain user emotional data and re-evaluate proposals based on this, making it possible to select innovative proposals with greater accuracy than conventional systems.

[0326] Prompt Sentence Examples

[0327] A user uses a web form to submit a new energy-efficient battery design or furniture made from recyclable materials. How does the server process these and ultimately evaluate them?

[0328] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0329] Step 1:

[0330] Users submit ideas through a web form or a dedicated application. For example, they enter an idea such as "Design a new, energy-efficient battery." The entered idea data is sent from the device to a server.

[0331] Input: Idea data entered by the user

[0332] Output: Idea data sent from the device to the server

[0333] What happens: A user enters an idea into a web form, clicks the submit button, and the data is sent to the server.

[0334] Step 2:

[0335] The server stores the received idea data in a database, including the idea text, submitter information, and submission date and time.

[0336] Input: Idea data sent from the device

[0337] Output: Idea data stored in a database

[0338] Specific operation: The server receives the idea data and records it as a new entry in the database.

[0339] Step 3:

[0340] The server retrieves raw ideas from the database, and performs preprocessing on the text data of the ideas, such as removing stop words, stemming, and extracting keywords.

[0341] Input: Raw idea data retrieved from the database

[0342] Output: Preprocessed text data

[0343] Specific operation: For example, stop words are removed from "new energy-efficient battery design" to "new energy-efficient battery design" and converted to the basic form.

[0344] Step 4:

[0345] The server converts the preprocessed text data of ideas into numerical vectors, and quantifies the importance and meaning of words using TF-IDF and Word2Vec.

[0346] Input: Preprocessed text data

[0347] Output: Numerical vector data

[0348] Specific operation: For example, the sentence "New energy-efficient high-battery design" is converted into a numerical vector [0.5, 0.7, 0.6, 0.8].

[0349] Step 5:

[0350] The server calculates the cosine similarity between the numeric vectors of all ideas, which produces a similarity score between proposals and identifies less relevant proposals.

[0351] Input: Numeric vector data

[0352] Output: Similarity score

[0353] Specific operation: For example, the cosine similarity between the numeric vectors of idea A and idea B is calculated to be 0.2.

[0354] Step 6:

[0355] The server extracts ideas with similarity scores below a certain threshold and stores these ideas in a database as "unique ideas."

[0356] Input: Similarity score

[0357] Output: Unique Idea List

[0358] Specific behavior: Ideas with a similarity score of 0.3 or less are extracted and saved in the database with a new label.

[0359] Step 7:

[0360] The server uses an emotion engine to acquire the user's emotional data, analyzes facial expressions and voice data when the user submits an idea, and quantifies the user's emotional state.

[0361] Input: User's facial expression data, voice data

[0362] Output: Quantified emotion data

[0363] Specific operation: The user's facial expressions and voice recorded via the webcam and microphone are analyzed, and the "excitement" state is quantified as 80%.

[0364] Step 8:

[0365] Based on the sentiment data obtained by the server, a recalculation is performed to influence the rating of the proposals. Suggestions with high positive sentiment are given higher ratings and prioritized.

[0366] Input: Quantified emotion data, unique idea list

[0367] Output: Re-evaluated unique idea list

[0368] Specific operation: The evaluation scores are calculated again and it is determined that "Idea A is first place, and Idea B is second place."

[0369] Step 9:

[0370] The server sends notifications to users and judges, including a summary of the unique ideas extracted and a request for confirmation and evaluation.

[0371] Input: Reevaluated Unique Ideas List

[0372] Output: Notification message

[0373] Specific operation: The judges will be notified of the "New Energy-Efficient High-Battery Design" and will receive an email requesting their evaluation.

[0374] (Application example 2)

[0375] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0376] Conventional proposal evaluation systems primarily analyze the content of proposals to assess their innovativeness and uniqueness. However, they ignore human factors such as the submitter's emotional state and motivation at the time of proposal, and therefore are unable to fully assess whether the submitter is truly motivated and whether the idea is innovative. This creates a risk of overlooking promising ideas, and it is necessary to improve the accuracy of evaluation results.

[0377] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for acquiring multiple submitted proposals, means for converting the proposals into numerical vectors, means for calculating the similarity of each proposal in a vector space, means for extracting proposals whose similarity is equal to or less than a certain threshold based on the calculated similarity, and means for analyzing the submitter's emotions and adjusting the evaluation of the proposal based on the emotion data. This enables a comprehensive evaluation that takes into account the submitter's emotional state in addition to evaluating the content of the proposal itself.

[0378] A "proposal" is an idea or plan for a new technology or service submitted to an innovation contest or idea pitch event.

[0379] A "numeric vector" is a set of coordinate values ​​in a multidimensional space used to analyze the proposal.

[0380] A "vector space" is a multidimensional analytical domain for representing proposals as numerical vectors.

[0381] "Similarity" is an index that evaluates the relevance of proposal content based on the distance and angle between numerical vectors.

[0382] The "threshold" is a boundary value set for determining similarity, below which a proposal is extracted as unique.

[0383] "Emotion data" is information about the emotional state of the person submitting the proposal, obtained by analyzing the person's facial expression, voice, etc.

[0384] "Evaluation" is the process of determining the innovativeness and potential value of a proposal based on its content and sentiment data.

[0385] "Analysis" involves examining submitted proposals and sentiment data in detail to clarify their nature and relevance.

[0386] In this invention, a system centered on a server, a terminal, and a user operates in cooperation with each other. A specific embodiment of the system will be described below.

[0387] 1. Collect and save ideas

[0388] Users submit new ideas through a terminal that includes a specialized application and allows users to input each idea by voice, which is converted into text using speech recognition technology and sent to a server.

[0389] 2. Preprocessing using natural language processing

[0390] The server preprocesses the received ideas using a natural language processing engine (e.g., spaCy or scikit-learn), which includes stopword removal, stemming, and keyword extraction.

[0391] 3. Conversion to a numeric vector

[0392] The server converts the preprocessed text data into a numerical vector using TF-IDF or Word2Vec, which allows the characteristics of each idea to be expressed numerically.

[0393] 4. Calculating similarity and extracting unique ideas

[0394] The server calculates the cosine similarity in the vector space to evaluate the similarity of each idea. Ideas with similarity below a certain threshold are extracted as unique ideas.

[0395] 5. Acquisition and Analysis of Emotion Data

[0396] The server uses a device equipped with a camera and microphone to capture the submitter's emotional data. It uses an emotion engine (e.g., OpenCV and TensorFlow) to analyze facial expressions and tone of voice, allowing it to quantify the submitter's emotional state when submitting an idea.

[0397] 6. Reassessment based on emotional data

[0398] The server uses the emotion data to re-evaluate unique ideas and adjust their ratings: if excitement or enthusiasm is high, the idea's score will increase.

[0399] 7. Final Evaluation and Notification

[0400] The server extracts the most innovative and unique ideas and sends the results through a notification system to notify factory managers and other stakeholders.

[0401] Specific use cases

[0402] Suppose an employee submits an idea for a new material handling system. The idea is converted into text using a robotic voice recognition system and sent to a server. After natural language processing, the text is converted into a numerical vector using TF-IDF. The proposal is then evaluated by calculating cosine similarity and performing sentiment analysis.

[0403] Prompt Sentence Examples

[0404] "Explain the process in which employees submit ideas for improving production technology efficiency and the robot analyzes them. Use an emotion engine to reflect the employee's emotional state in the evaluation of the idea."

[0405] The present invention realizes highly accurate proposal evaluation by taking into consideration not only the content of the proposal but also the feelings of the submitter.

[0406] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0407] Step 1:

[0408] The device converts the user's input ideas into text data using voice recognition technology. Specifically, the device receives voice input through a built-in microphone, and the voice recognition engine converts this into text data. The input is voice data, and the output is text data.

[0409] Step 2:

[0410] The server receives the text data sent from the terminal. At this stage, the idea submitted by the user is saved in text format on the server. The input is text data, and the output is idea data stored in a database.

[0411] Step 3:

[0412] The server uses a natural language processing engine to preprocess the text data. Specifically, it performs stop word removal, stemming, and keyword extraction. The input is the text data, and the output is the preprocessed text data.

[0413] Step 4:

[0414] The server uses TF-IDF or Word2Vec to convert the preprocessed text data into a numerical vector, where the input is the preprocessed text data and the output is a numerical vector.

[0415] Step 5:

[0416] The server calculates the cosine similarity between the numeric vectors. Specifically, it calculates the similarity score using the inner product of each numeric vector and the vector norm. The input is the numeric vector, and the output is the similarity score.

[0417] Step 6:

[0418] The server extracts ideas whose similarity scores are below a certain threshold, where the input is the similarity score and the output is unique ideas.

[0419] Step 7:

[0420] The device uses a camera and microphone to analyze the user's facial expressions and voice. It uses an emotion engine to acquire the user's emotional data. The input is facial expression data and voice data, and the output is emotional data.

[0421] Step 8:

[0422] The server adjusts the evaluation of ideas based on the emotional data. Specifically, it increases the evaluation score of ideas with positive emotions. The inputs are the emotional data and unique idea data, and the output is the re-evaluated idea data.

[0423] Step 9:

[0424] The server notifies the factory manager of the final evaluation result of the unique idea. The input is the re-evaluated idea data, and the output is a notification message.

[0425] In this way, the system takes into account both the content of the proposal and the emotional state of the submitter, allowing it to efficiently extract and evaluate the most innovative ideas.

[0426] 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.

[0427] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0428] 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.

[0429] [Second embodiment]

[0430] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0431] 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.

[0432] 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).

[0433] 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.

[0434] 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.

[0435] 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).

[0436] 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.

[0437] 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.

[0438] 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.

[0439] 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.

[0440] 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.

[0441] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0442] This invention relates to a system that efficiently analyzes multiple proposals submitted in innovation contests and idea pitch events and automatically extracts innovative proposals. This system includes the following processing steps, and can consistently perform proposal submission, processing, and evaluation.

[0443] Program processing procedure

[0444] 1. Idea collection

[0445] Users submit ideas to innovation contests and idea pitch events via web forms or dedicated applications.

[0446] 2. Save your ideas in a database

[0447] The server receives the submitted ideas and stores them in a database. The stored data includes the idea text, submitter information, and submission date and time.

[0448] 3. Preprocessing ideas using natural language processing

[0449] The server retrieves unprocessed ideas from the database and performs the following natural language processing on the retrieved ideas.

[0450] Stop word removal: remove common meaningless words (e.g. "no", "wa", "ni", etc.).

[0451] Stemming: Extracting the base form of a word. (Example: "hashiru" → "hashiru")

[0452] Keyword extraction: Extract important words and phrases.

[0453] 4. Converting ideas into numerical vectors

[0454] The server converts the preprocessed idea text data into a numerical vector using the following techniques:

[0455] TF-IDF (Term Frequency-Inverse Document Frequency): Evaluates and quantifies the importance of each word.

[0456] Word2Vec: Generates numeric vectors based on word meanings.

[0457] 5. Similarity calculation in vector space

[0458] The server calculates the cosine similarity between the numeric vectors of all ideas, which allows it to calculate a similarity score between proposals and identify less relevant proposals.

[0459] 6. Extracting unique ideas

[0460] The server extracts ideas with similarity scores below a certain threshold. These are considered unique ideas that are significantly different from other proposals and are stored in a database as "unique ideas."

[0461] 7. Notices and Confirmations

[0462] The server notifies the judges and users, prompting them to review and evaluate the extracted unique ideas.

[0463] Specific examples

[0464] Example: Idea submission and processing

[0465] 1. A user submits two ideas using a web form, for example, "a new energy-efficient battery design" and "furniture made from recyclable materials."

[0466] 2. The server stores these ideas in a database and marks them as pending.

[0467] 3. The server performs natural language processing to remove stop words, stem, and extract key keywords from each idea.

[0468] Idea A Keywords: ["energy", "efficiency", "battery", "design"]

[0469] Idea B Keywords: ["recyclable", "material", "furniture"]

[0470] 4. The server generates a numeric vector based on the keyword.

[0471] Idea A's vector: [0.5, 0.7, 0.6, 0.8]

[0472] Idea B vector: [0.1, 0.3, 0.9, 0.4]

[0473] 5. The server calculates the cosine similarity and finds that the similarity score between these ideas is 0.2. Based on this low score, Idea A and Idea B are extracted as unique ideas.

[0474] 6. The server notifies the judges and prompts them to review and evaluate the extracted ideas.

[0475] This system makes it possible to efficiently extract innovative and unique ideas from a large number of proposals.

[0476] The processing flow will be explained below.

[0477] Step 1:

[0478] A user enters their idea in a web form or a dedicated application and clicks the submit button. The idea is entered in text format.

[0479] Step 2:

[0480] The terminal transmits the idea input by the user to the server, and the transmitted data includes the idea text, the user's identification information, the submission date and time, etc.

[0481] Step 3:

[0482] The server stores the received idea data in the database, and the stored data is marked with an unprocessed flag for the idea.

[0483] Step 4:

[0484] The server retrieves ideas with unprocessed flags from the database, and the retrieved ideas are subjected to natural language processing.

[0485] Step 5:

[0486] The server removes stop words from the idea text, which in the case of Japanese is a process to remove common words such as "no", "wa", and "ni".

[0487] Step 6:

[0488] The server performs stemming and converts words to their base form, for example, "run" and "ran" are converted to "run".

[0489] Step 7:

[0490] The server performs keyword extraction and lists important words and phrases, which are later converted into numerical vectors.

[0491] Step 8:

[0492] The server takes the preprocessed text and converts it into a numerical vector, using techniques such as TF-IDF or Word2Vec.

[0493] Step 9:

[0494] The server calculates the cosine similarity between the generated numeric vectors for all ideas. This similarity score evaluates the relative relevance of the proposals.

[0495] Step 10:

[0496] The server evaluates the similarity scores and extracts ideas with scores below a set threshold, which is used to identify ideas that are unique and different from other proposals.

[0497] Step 11:

[0498] The server stores the extracted unique ideas in a database and sets a special flag so that they are given priority access in subsequent evaluation processes.

[0499] Step 12:

[0500] The server sends notifications to users and judges, including a summary of the unique ideas extracted and a request for confirmation and evaluation.

[0501] Example 1

[0502] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0503] Conventional proposal extraction systems have had difficulty efficiently extracting innovative and unique proposals from a large number of proposals. Furthermore, the accuracy of proposal preprocessing and similarity calculations was insufficient, making it impossible to guarantee the quality of the extracted proposals. As a result, proposal evaluation took a long time, increasing the burden on judges and users.

[0504] 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.

[0505] In this invention, the server includes means for acquiring a plurality of submitted proposals, means for storing the acquired proposals in a database, means for acquiring unprocessed proposals from the database, means for preprocessing text data of the proposals using natural language processing, means for converting the preprocessed proposals into numerical vectors, means for calculating the similarity of each proposal in a vector space, means for extracting proposals whose similarity is equal to or less than a certain threshold based on the calculated similarity, and means for notifying the extracted proposals. This makes it possible to efficiently extract innovative and unique proposals from a large number of proposals.

[0506] "Proposal" means an idea or outline submitted by a User in an innovation contest or idea pitch event.

[0507] "Database" refers to a collection of information for storing and managing submitted proposals and related information (such as submitter information, submission date and time, etc.).

[0508] "Natural language processing" is a general term for technology that analyzes text data, understands its meaning, and extracts necessary information.

[0509] A "numeric vector" is a mathematical representation of text data or proposal content in a format that allows for calculations within a vector space.

[0510] "Cosine similarity" is a calculation method for evaluating the similarity between two numerical vectors, and uses the angle between the vectors.

[0511] "TF-IDF" stands for "Term Frequency-Inverse Document Frequency" and is a method for calculating and quantifying the importance of each word.

[0512] "Word2Vec" is a technology for representing the meaning of words as numerical vectors, and is a model that trains words that are semantically similar to each other so that they have similar vectors.

[0513] "Notification" refers to the action of sending information so that judges and users are aware of the proposal.

[0514] The present invention relates to a system that efficiently analyzes multiple proposals submitted in innovation contests and idea pitch events and automatically extracts innovative proposals. This system operates using the following hardware and software.

[0515] Hardware and software used

[0516] Server: A computer system for database management, natural language processing, numeric vector transformation, similarity calculation, and notification sending.

[0517] Database: An information management system for storing proposal data, user information, and similarity scores.

[0518] Natural language processing libraries: For example, use Python's nlp library to analyze text data.

[0519] Numerical vector transformation libraries: For example, use the scikit-learn library to perform TF-IDF and Word2Vec transformations.

[0520] Similarity calculation library: For example, use the scipy library to calculate cosine similarity.

[0521] Notification Sending API: API for sending emails and push notifications.

[0522] Example of a system

[0523] Example: Idea submission and processing

[0524] 1. A user submits an idea for a "new energy-efficient battery design" using a web form.

[0525] 2. The server receives the idea and stores it in the database in the format "Idea: New energy-efficient battery design, Submitted by: Username, Submitted date: 2023-10-01".

[0526] 3. The server retrieves raw ideas from the database and performs natural language processing, such as removing "no," "wa," and "ni" from the idea text, stemming "hashiru" to "so," and extracting keywords such as "energy," "efficiency," "battery," and "design."

[0527] 4. The server calculates the TF-IDF based on these keywords and generates a numerical vector, for example, "energy" is 0.7, "efficiency" is 0.8, "battery" is 0.9, and "design" is 1.0.

[0528] 5. The server calculates the cosine similarity between all ideas and obtains a similarity score (e.g., 0.2).

[0529] 6. The server extracts ideas with a similarity score below a threshold (e.g., 0.3) as unique ideas and stores them in the database. For example, it flags them as "unique ideas" using an SQL UPDATE statement.

[0530] 7. The server sends a notification email to the judges, prompting them to confirm and evaluate the unique ideas.

[0531] Example prompts for generative AI models

[0532] Here are some example prompts you can enter into your generative AI model:

[0533] "Please extract important keywords from the following text: 'New energy-efficient battery design'"

[0534] The present invention makes it possible to efficiently extract innovative and unique proposals from a large number of proposals, thereby improving the efficiency of the review and the quality of the evaluation.

[0535] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0536] Step 1:

[0537] A user submits an idea. Using a web form, they enter the title and body of the idea and click the "Submit" button. The input is the title and body of the idea, and the output is the idea sent to the server.

[0538] Step 2:

[0539] The server receives the submitted idea. This temporarily stores the submitted idea data (title, text, submitter information, submission date and time). The input is the idea data sent by the user, and the output is the data stored in the server's temporary memory.

[0540] Step 3:

[0541] The server saves the received idea data in the database. The server establishes a database connection and uses the SQL INSERT statement to store the idea data in persistent storage. As input, there is idea data stored in temporary memory, and as output, the idea data is saved in the database.

[0542] Step 4:

[0543] The server retrieves unprocessed ideas from the database. It uses a SQL SELECT statement to extract idea data that is marked as unprocessed. As input, it has idea data that is flagged as unprocessed, and as output, it extracts that data to the server.

[0544] Step 5:

[0545] The server performs natural language processing (NLP) to preprocess the idea data. It uses Python's NLP library to remove stop words, perform stemming, and extract keywords. The extracted idea data is input, and preprocessed idea data is generated as output. Specifically, it removes words such as "no," "wa," and "ni" from the main body of the idea, converts "hashiru" to "so," and extracts keywords such as "energy," "efficiency," "battery," and "design."

[0546] Step 6:

[0547] The server converts the preprocessed idea data into a numerical vector. Using Python's scikit-learn library, the conversion is performed using TF-IDF and Word2Vec. The preprocessed idea data is input, and a numerical vector is generated as output. Specifically, the importance of each keyword is evaluated and quantified.

[0548] Step 7:

[0549] The server calculates the cosine similarity between all idea vectors. The similarity calculation is performed using the scipy library. The input is a set of numerical vectors, and the output is a similarity score between each vector. Specifically, the angle between the vectors is calculated and the result is scored.

[0550] Step 8:

[0551] The server extracts unique ideas based on the similarity score. The server stores ideas with a similarity score below a certain threshold as "unique ideas" in a database. The input is the similarity score, and the output is data flagged as a unique idea.

[0552] Step 9:

[0553] The server sends notifications to judges and users. The server uses an email sending API to notify unique ideas that require evaluation. The input is the data of the unique idea, and the output is a notification email. Specifically, the body of the notification email contains the title of the idea and an evaluation link.

[0554] (Application example 1)

[0555] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0556] Proposing suggestions to improve work efficiency on-site requires a lot of time and effort. In particular, the process of analyzing information collected from the factory floor, extracting useful suggestions, and quickly communicating the results to workers is complex, which can delay responses when immediate action is required. In addition, traditional methods often include many overlapping or similar suggestions, which can result in useful suggestions being overlooked.

[0557] 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.

[0558] In this invention, the server includes a means for acquiring a plurality of submitted proposals, a means for converting the proposals into a numerical vector, a means for calculating similarity, a means for extracting proposals whose similarity is equal to or less than a certain threshold, and a means for using a smart device to present the extracted proposals. This makes it possible to quickly identify useful proposals for supporting improved work efficiency at the factory site and quickly notify workers.

[0559] The "means for acquiring a plurality of submitted proposals" refers to a means by which workers or engineers on the factory floor use smart devices to input proposals and transmit them to a server.

[0560] The "means for converting into a numerical vector" is a means for converting the proposed text data into a numerical vector using a specific algorithm (TF-IDF or Word2Vec).

[0561] The "means for calculating similarity" is a means for performing calculations to evaluate the similarity between proposals using the converted numerical vectors.

[0562] The "means for extracting proposals whose similarity is equal to or less than a certain threshold" is a means for identifying and extracting unique proposals that are significantly different from other proposals, based on the similarity scores between proposals.

[0563] The "means of using a smart device to present extracted suggestions" refers to a means of notifying and displaying the extracted unique suggestions to workers or engineers in real time through a smart device such as smart glasses.

[0564] The present invention is a proposed analysis system aimed at improving work efficiency at a factory site. A specific embodiment of this system is described below.

[0565] First, workers or engineers use smart glasses to voice-input improvements or new ideas they discover on-site. The voice-input content is converted into text data by the voice recognition software in the smart glasses. This text data is then sent to a server via a network. This process uses smart glasses (e.g., Google Glass) and their built-in voice recognition software (e.g., Google Speech-to-Text API).

[0566] The server stores the received suggestions in a database and then performs the following natural language processing: The server removes common meaningless words (stop words) from the suggested text and converts words to their root forms (stemming). This process is performed using the NLTK library. After that, the text data is converted into a numerical vector using the TF-IDF (Term Frequency-Inverse Document Frequency) or Word2Vec algorithm to extract important keywords. This is performed using the Scikit-learn library.

[0567] Next, the server calculates the cosine similarity between the vectors and calculates a similarity score for each proposal. It identifies proposals with a similarity score below a certain threshold (unique proposals) and extracts them. The Scikit-learn library is also used to calculate the similarity.

[0568] The extracted unique suggestions are sent to workers and engineers using smart glasses. The notification includes an overview of the suggestion and its evaluation, allowing workers to immediately check and evaluate the content of the suggestion. This enables quick responses on-site and is expected to lead to efficient work improvements.

[0569] As a specific use case, field worker A makes a suggestion by voice, "Design a new, energy-efficient battery." This suggestion is sent to the server, where it undergoes natural language processing and similarity calculations are performed. Similarly, worker B makes a suggestion, "Furniture made from recyclable materials." The server evaluates the similarity between these suggestions and identifies both as unique suggestions. This allows workers to instantly receive useful suggestions and incorporate them into their improvement activities.

[0570] An example of an input prompt for the generative AI model is as follows:

[0571] Next, enter the idea you want to collect. For example, enter a brief description of your idea, such as "new energy-efficient battery designs" or "furniture made from recyclable materials."

[0572] By using the above-described means, the present invention improves the efficiency of a series of processes from collection of proposals to analysis and notification at the factory site, thereby contributing to improved work efficiency.

[0573] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0574] Step 1:

[0575] The user uses the smart glasses to speak suggestions, which are converted into text by voice recognition software within the glasses, and then sent to the server.

[0576] Input: User's voice suggestions

[0577] Output: Proposals converted to text data

[0578] How it works: The smart glasses record the voice suggestions and convert them into text using voice recognition software, which then sends the text data over the network to a server.

[0579] Step 2:

[0580] The server stores the received proposal in a database. Data including details of the proposal (text, submitter information, submission date and time, etc.) is recorded in the database.

[0581] Input: Text data

[0582] Output: Proposal data stored in a database

[0583] What happens: The server receives the text data and stores it as a new record in the database. All necessary metadata is also saved at this stage.

[0584] Step 3:

[0585] The server retrieves the raw stored suggestions and performs natural language processing on them, specifically stop word removal, stemming, and keyword extraction.

[0586] Input: Proposal data retrieved from the database

[0587] Output: Preprocessed text data

[0588] Specific operation: The server uses the NLTK library to remove stop words and perform stemming, and performs keyword extraction using algorithms such as TF-IDF and Word2Vec.

[0589] Step 4:

[0590] The server converts the preprocessed text data of the proposals into a numerical vector using TF-IDF or Word2Vec.

[0591] Input: Preprocessed text data

[0592] Output: Numeric vector

[0593] Specific operation: The server converts text data into a numerical vector using TF-IDF or Word2Vec from the Scikit-learn library.

[0594] Step 5:

[0595] The server calculates the cosine similarity between all numeric vectors, which then produces a similarity score between the proposals.

[0596] Input: Numeric vector

[0597] Output: Similarity score

[0598] Specific operation: The server uses Scikit-learn's cosine_similarity method to calculate the cosine similarity between numeric vectors and obtain a similarity score.

[0599] Step 6:

[0600] The server extracts proposals with similarity scores below a certain threshold and identifies them as unique proposals.

[0601] Input: Similarity score

[0602] Output: A list of unique suggestions

[0603] Specific operation: The server evaluates the similarity score based on the set threshold and extracts a list of unique proposals that are significantly different from other proposals.

[0604] Step 7:

[0605] The server notifies the user of the extracted unique suggestions through the smart glasses, along with a summary of the suggestions and their ratings.

[0606] Input: A list of unique suggestions

[0607] Output: Notification of the suggestion displayed on the smart glasses

[0608] Specific operation: The server composes the extracted unique ideas into a notification message and sends it to the smart glasses via the network. The smart glasses then display the content of the suggestion to the user.

[0609] Through the above processing steps, the present invention makes it possible to quickly identify useful suggestions for improving work efficiency on the factory floor and notify workers of them.

[0610] 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.

[0611] The present invention achieves even more accurate evaluation by combining a system that analyzes multiple proposals submitted to innovation contests and idea pitch events and automatically extracts innovative proposals with an emotion engine that recognizes the user's emotions.

[0612] Program processing procedure

[0613] 1. Idea collection

[0614] Users submit ideas to innovation contests and idea pitch events via web forms or dedicated applications.

[0615] 2. Save your ideas in a database

[0616] The server receives the submitted ideas and stores them in a database. The stored data includes the idea text, submitter information, and submission date and time.

[0617] 3. Preprocessing ideas using natural language processing

[0618] The server retrieves unprocessed ideas from the database and performs the following natural language processing on the retrieved ideas.

[0619] Stop word removal: remove common meaningless words (e.g. "no", "wa", "ni", etc.).

[0620] Stemming: Extracting the base form of a word. (Example: "hashiru" → "hashiru")

[0621] Keyword extraction: Extract important words and phrases.

[0622] 4. Converting ideas into numerical vectors

[0623] The server converts the preprocessed idea text data into a numerical vector using the following techniques:

[0624] TF-IDF (Term Frequency-Inverse Document Frequency): Evaluates and quantifies the importance of each word.

[0625] Word2Vec: Generates numeric vectors based on word meanings.

[0626] 5. Similarity calculation in vector space

[0627] The server calculates the cosine similarity between the numeric vectors of all ideas, which allows it to calculate a similarity score between proposals and identify less relevant proposals.

[0628] 6. Extracting unique ideas

[0629] The server extracts ideas with similarity scores below a certain threshold. These are considered unique ideas that are significantly different from other proposals and are stored in a database as "unique ideas."

[0630] 7. Leveraging Emotional Engines

[0631] The server uses an emotion engine to acquire the user's emotional data, analyzing facial expressions and voice data when the user submits an idea, and quantifying the user's emotional state.

[0632] 8. Reflecting Emotional Data in Evaluation

[0633] The server influences the evaluation of the proposal based on the emotional data it acquires. For example, if the user has very positive emotions toward a proposal, it will increase the innovativeness rating of that proposal.

[0634] 9. Reevaluate unique ideas

[0635] The server performs re-evaluation using the emotion engine. Based on the emotion data, it recalculates the evaluation of unique ideas that have already been extracted and prioritizes them.

[0636] 10. Notices and Confirmations

[0637] The server sends notifications to users and judges, including a summary of the unique ideas extracted and a request for confirmation and evaluation.

[0638] Specific examples

[0639] Example: Idea submission and processing

[0640] 1. A user submits two ideas using a web form, for example, "a new energy-efficient battery design" and "furniture made from recyclable materials."

[0641] 2. The device sends these ideas to the server.

[0642] 3. The server stores these ideas in a database and marks them as pending.

[0643] 4. The server performs natural language processing to remove stop words, stem, and extract key keywords from each idea.

[0644] Idea A Keywords: ["energy", "efficiency", "battery", "design"]

[0645] Idea B Keywords: ["recyclable", "material", "furniture"]

[0646] 5. The server generates a numeric vector based on the keyword.

[0647] Idea A's vector: [0.5, 0.7, 0.6, 0.8]

[0648] Idea B vector: [0.1, 0.3, 0.9, 0.4]

[0649] 6. The server calculates the cosine similarity and finds that the similarity score between these ideas is 0.2. Based on this low score, Idea A and Idea B are extracted as unique ideas.

[0650] 7. The server uses an emotion engine to extract emotional data from facial expressions and voice when users submit ideas. For example, if a user is very excited, the idea will be rated higher.

[0651] 8. The server re-evaluates and prioritizes ideas based on sentiment data.

[0652] 9. The server notifies the judges and prompts them to review and evaluate the extracted ideas.

[0653] This system efficiently extracts innovative and unique ideas from a large number of proposals, and by evaluating them based on user emotional data, it is possible to achieve highly accurate evaluations.

[0654] The processing flow will be explained below.

[0655] Step 1:

[0656] A user enters their idea in a web form or a dedicated application and clicks the submit button. The idea is entered in text format.

[0657] Step 2:

[0658] The terminal transmits the idea input by the user to the server, and the transmitted data includes the idea text, the user's identification information, the submission date and time, etc.

[0659] Step 3:

[0660] The server stores the received idea data in the database, and the stored data is marked with an unprocessed flag for the idea.

[0661] Step 4:

[0662] The server retrieves ideas with unprocessed flags from the database, and the retrieved ideas are subjected to natural language processing.

[0663] Step 5:

[0664] The server removes stop words from the idea text, which in the case of Japanese is a process to remove common words such as "no", "wa", and "ni".

[0665] Step 6:

[0666] The server performs stemming and converts words to their base form, for example, "run" and "ran" are converted to "run".

[0667] Step 7:

[0668] The server performs keyword extraction and lists important words and phrases, which are later converted into numerical vectors.

[0669] Step 8:

[0670] The server takes the preprocessed text and converts it into a numerical vector, using techniques such as TF-IDF or Word2Vec.

[0671] Step 9:

[0672] The server calculates the cosine similarity between the generated numeric vectors for all ideas. This similarity score evaluates the relative relevance of the proposals.

[0673] Step 10:

[0674] The server evaluates the similarity scores and extracts ideas with scores below a set threshold, which is used to identify ideas that are unique and different from other proposals.

[0675] Step 11:

[0676] The server stores the extracted unique ideas in a database and sets a special flag so that they are given priority access in subsequent evaluation processes.

[0677] Step 12:

[0678] The server uses an emotion engine to acquire emotional data from the user's facial expressions and voice when submitting an idea. The emotional data is analyzed in real time and quantified to represent the user's emotional state (e.g., excitement, joy, surprise, etc.).

[0679] Step 13:

[0680] The server reflects the emotional data in the idea evaluation. Positive emotional data positively influences the evaluation score of the idea, while negative emotional data is used as an indicator to reconsider the evaluation.

[0681] Step 14:

[0682] The server uses the results of the emotion data to re-evaluate the unique ideas that have already been extracted. As a result of the re-evaluation, the priorities may change.

[0683] Step 15:

[0684] The server sends a notification to the user and judges. The notification includes a summary of the unique ideas extracted, as well as the results of a re-evaluation that takes into account the emotional data. The judges then make their final confirmation and evaluation based on this information.

[0685] The following is an explanation based on a specific example.

[0686] Example: Submitting and processing an idea

[0687] Step 1 and Step 2:

[0688] Users submit two ideas through a web form: "A new energy-efficient battery design" and "Furniture made from recyclable materials." The device then sends the ideas to the server.

[0689] Step 3 to Step 7:

[0690] The server stores these ideas in a database and performs natural language processing on each idea, removing stop words, stemming, and extracting important keywords.

[0691] For example, create a list of keywords such as "energy efficient battery design" and "recyclable materials furniture."

[0692] Steps 8 to 10:

[0693] The server converts keywords into numerical vectors and calculates the cosine similarity. Ideas with low similarity scores are extracted as unique ideas and stored in a database.

[0694] Step 11 to Step 14:

[0695] The server uses an emotion engine to obtain user emotion data and reflect it in the idea evaluation. For example, if a user is very excited when submitting a "new energy-efficient battery design," the evaluation score of the idea will be increased.

[0696] Re-evaluation is performed based on the emotional data, and the priority of the extracted unique ideas is recalculated.

[0697] Step 15:

[0698] The server notifies the judges and users and provides an overview of the unique ideas extracted and evaluation results that take into account the emotional data. The judges then make their final confirmation and evaluation based on this information.

[0699] This makes it possible to accurately evaluate and extract ideas that incorporate emotional data.

[0700] Example 2

[0701] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0702] Conventional idea evaluation systems have difficulty efficiently extracting innovative and unique proposals from a large number of proposals. Furthermore, they lack user emotional data, resulting in low evaluation accuracy. A system that can solve these issues and provide more accurate and efficient evaluations is needed.

[0703] 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.

[0704] In this invention, the server includes means for acquiring a plurality of submitted proposals, means for converting the proposals into numerical vectors, means for calculating the similarity of each proposal in a vector space, means for extracting proposals whose similarity is equal to or less than a certain threshold based on the calculated similarity, means for acquiring user emotion data using an emotion engine, and means for recalculating proposal evaluations based on the acquired emotion data. This makes it possible to efficiently extract innovative and unique proposals from a large number of proposals and to enable highly accurate evaluations that reflect the user's emotion data.

[0705] "Submitted Proposals" means ideas or proposals submitted by a User to the System through a contest or event.

[0706] "Means for obtaining" refers to the process or component by which the server receives and stores user input data.

[0707] "Means for converting to a numeric vector" refers to a process or algorithm for converting text data into a numeric vector format using natural language processing techniques.

[0708] "Means for calculating the similarity of each proposal in vector space" refers to a process or algorithm for calculating, for example, the cosine similarity between proposals converted into numerical vectors.

[0709] "Means for extracting proposals with similarity below a certain threshold" refers to a process or algorithm for selecting unique proposals that are significantly different from other proposals based on a calculated similarity score.

[0710] "Means for acquiring user emotional data using an emotion engine" refers to a process or component that uses technologies such as facial expression analysis and voice analysis to detect the user's emotional state and acquire it as data.

[0711] "Means for recalculating the evaluation of a proposal based on acquired emotional data" refers to a process or algorithm for correcting or recalculating the evaluation of the innovativeness or uniqueness of a proposal based on the emotional data.

[0712] This invention relates to a system that analyzes multiple proposals submitted to innovation contests and idea pitch events and automatically extracts innovative proposals. This system achieves more accurate evaluation by combining it with an emotion engine that recognizes the user's emotions.

[0713] The system uses the following hardware and software:

[0714] Server: Receives, stores, analyzes, and evaluates proposal data.

[0715] Terminal: Runs a web form or dedicated application that allows users to submit ideas.

[0716] Emotion engine: Facial expression and voice analysis technology to recognize user emotions (e.g., facial expression recognition software, voice recognition software).

[0717] System hardware and software configuration

[0718] 1. A means of retrieving multiple submitted proposals

[0719] Users submit ideas through a web form or a dedicated application, for example, by entering an idea such as "Design a new, energy-efficient battery."

[0720] 2. A means of converting proposals into numerical vectors

[0721] The server uses natural language processing technology to convert the text data of submitted ideas into numerical vectors, specifically using TF-IDF and Word2Vec to quantify the importance and meaning of words.

[0722] 3. A means to calculate the similarity of each proposal in the vector space

[0723] The server calculates the cosine similarity between the proposals converted into numerical vectors, which allows us to evaluate the relevance between the proposals.

[0724] 4. A method for extracting proposals whose similarity is below a certain threshold

[0725] The server extracts unique proposals that are significantly different from other proposals based on the calculated similarity scores, thus identifying innovative and unique proposals.

[0726] 5. Means of acquiring user emotion data using an emotion engine

[0727] The server uses an emotion engine to analyze facial expressions and voice data when a user submits a suggestion and quantify their emotional state. For example, a user who is very excited might be recorded as 80%.

[0728] 6. A method for recalculating the evaluation of proposals based on acquired emotional data

[0729] The server reflects the sentiment data and recalculates the evaluation of the proposal's innovativeness and uniqueness, so that the user's sentiment influences the proposal's final evaluation.

[0730] As a concrete example, consider a case where a user submits two ideas using a web form, say "designing a new energy-efficient battery" and "furniture made from recyclable materials." The device sends these ideas to a server, which stores them in a database and marks them as unprocessed. The server then performs natural language processing (NLP) to remove stop words from each idea, perform stemming, and extract key keywords. The server then generates a numerical vector based on the keywords and calculates the cosine similarity to calculate a similarity score between the proposals.

[0731] A particularly important step is to use an emotion engine to obtain user emotional data and re-evaluate proposals based on this, making it possible to select innovative proposals with greater accuracy than conventional systems.

[0732] Prompt Sentence Examples

[0733] A user uses a web form to submit a new energy-efficient battery design or furniture made from recyclable materials. How does the server process these and ultimately evaluate them?

[0734] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0735] Step 1:

[0736] Users submit ideas through a web form or a dedicated application. For example, they enter an idea such as "Design a new, energy-efficient battery." The entered idea data is sent from the device to a server.

[0737] Input: Idea data entered by the user

[0738] Output: Idea data sent from the device to the server

[0739] What happens: A user enters an idea into a web form, clicks the submit button, and the data is sent to the server.

[0740] Step 2:

[0741] The server stores the received idea data in a database, including the idea text, submitter information, and submission date and time.

[0742] Input: Idea data sent from the device

[0743] Output: Idea data stored in a database

[0744] Specific operation: The server receives the idea data and records it as a new entry in the database.

[0745] Step 3:

[0746] The server retrieves raw ideas from the database, and performs preprocessing on the text data of the ideas, such as removing stop words, stemming, and extracting keywords.

[0747] Input: Raw idea data retrieved from the database

[0748] Output: Preprocessed text data

[0749] Specific operation: For example, stop words are removed from "new energy-efficient battery design" to "new energy-efficient battery design" and converted to the basic form.

[0750] Step 4:

[0751] The server converts the preprocessed text data of ideas into numerical vectors, and quantifies the importance and meaning of words using TF-IDF and Word2Vec.

[0752] Input: Preprocessed text data

[0753] Output: Numerical vector data

[0754] Specific operation: For example, the sentence "New energy-efficient high-battery design" is converted into a numerical vector [0.5, 0.7, 0.6, 0.8].

[0755] Step 5:

[0756] The server calculates the cosine similarity between the numeric vectors of all ideas, which produces a similarity score between proposals and identifies less relevant proposals.

[0757] Input: Numeric vector data

[0758] Output: Similarity score

[0759] Specific operation: For example, the cosine similarity between the numeric vectors of idea A and idea B is calculated to be 0.2.

[0760] Step 6:

[0761] The server extracts ideas with similarity scores below a certain threshold and stores these ideas in a database as "unique ideas."

[0762] Input: Similarity score

[0763] Output: Unique Idea List

[0764] Specific behavior: Ideas with a similarity score of 0.3 or less are extracted and saved in the database with a new label.

[0765] Step 7:

[0766] The server uses an emotion engine to acquire the user's emotional data, analyzes facial expressions and voice data when the user submits an idea, and quantifies the user's emotional state.

[0767] Input: User's facial expression data, voice data

[0768] Output: Quantified emotion data

[0769] Specific operation: The user's facial expressions and voice recorded via the webcam and microphone are analyzed, and the "excitement" state is quantified as 80%.

[0770] Step 8:

[0771] Based on the sentiment data obtained by the server, a recalculation is performed to influence the rating of the proposals. Suggestions with high positive sentiment are given higher ratings and prioritized.

[0772] Input: Quantified emotion data, unique idea list

[0773] Output: Re-evaluated unique idea list

[0774] Specific operation: The evaluation scores are calculated again and it is determined that "Idea A is first place, and Idea B is second place."

[0775] Step 9:

[0776] The server sends notifications to users and judges, including a summary of the unique ideas extracted and a request for confirmation and evaluation.

[0777] Input: Reevaluated Unique Ideas List

[0778] Output: Notification message

[0779] Specific operation: The judges will be notified of the "New Energy-Efficient High-Battery Design" and will receive an email requesting their evaluation.

[0780] (Application example 2)

[0781] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0782] Conventional proposal evaluation systems primarily analyze the content of proposals to assess their innovativeness and uniqueness. However, they ignore human factors such as the submitter's emotional state and motivation at the time of proposal, and therefore are unable to fully assess whether the submitter is truly motivated and whether the idea is innovative. This creates a risk of overlooking promising ideas, and it is necessary to improve the accuracy of evaluation results.

[0783] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for acquiring multiple submitted proposals, means for converting the proposals into numerical vectors, means for calculating the similarity of each proposal in a vector space, means for extracting proposals whose similarity is equal to or less than a certain threshold based on the calculated similarity, and means for analyzing the submitter's emotions and adjusting the evaluation of the proposal based on the emotion data. This enables a comprehensive evaluation that takes into account the submitter's emotional state in addition to evaluating the content of the proposal itself.

[0784] A "proposal" is an idea or plan for a new technology or service submitted to an innovation contest or idea pitch event.

[0785] A "numeric vector" is a set of coordinate values ​​in a multidimensional space used to analyze the proposal.

[0786] A "vector space" is a multidimensional analytical domain for representing proposals as numerical vectors.

[0787] "Similarity" is an index that evaluates the relevance of proposal content based on the distance and angle between numerical vectors.

[0788] The "threshold" is a boundary value set for determining similarity, below which a proposal is extracted as unique.

[0789] "Emotion data" is information about the emotional state of the person submitting the proposal, obtained by analyzing the person's facial expression, voice, etc.

[0790] "Evaluation" is the process of determining the innovativeness and potential value of a proposal based on its content and sentiment data.

[0791] "Analysis" involves examining submitted proposals and sentiment data in detail to clarify their nature and relevance.

[0792] In this invention, a system centered on a server, a terminal, and a user operates in cooperation with each other. A specific embodiment of the system will be described below.

[0793] 1. Collect and save ideas

[0794] Users submit new ideas through a terminal that includes a specialized application and allows users to input each idea by voice, which is converted into text using speech recognition technology and sent to a server.

[0795] 2. Preprocessing using natural language processing

[0796] The server preprocesses the received ideas using a natural language processing engine (e.g., spaCy or scikit-learn), which includes stopword removal, stemming, and keyword extraction.

[0797] 3. Conversion to a numeric vector

[0798] The server converts the preprocessed text data into a numerical vector using TF-IDF or Word2Vec, which allows the characteristics of each idea to be expressed numerically.

[0799] 4. Calculating similarity and extracting unique ideas

[0800] The server calculates the cosine similarity in the vector space to evaluate the similarity of each idea. Ideas with similarity below a certain threshold are extracted as unique ideas.

[0801] 5. Acquisition and Analysis of Emotion Data

[0802] The server uses a device equipped with a camera and microphone to capture the submitter's emotional data. It uses an emotion engine (e.g., OpenCV and TensorFlow) to analyze facial expressions and tone of voice, allowing it to quantify the submitter's emotional state when submitting an idea.

[0803] 6. Reassessment based on emotional data

[0804] The server uses the emotion data to re-evaluate unique ideas and adjust their ratings: if excitement or enthusiasm is high, the idea's score will increase.

[0805] 7. Final Evaluation and Notification

[0806] The server extracts the most innovative and unique ideas and sends the results through a notification system to notify factory managers and other stakeholders.

[0807] Specific use cases

[0808] Suppose an employee submits an idea for a new material handling system. The idea is converted into text using a robotic voice recognition system and sent to a server. After natural language processing, the text is converted into a numerical vector using TF-IDF. The proposal is then evaluated by calculating cosine similarity and performing sentiment analysis.

[0809] Prompt Sentence Examples

[0810] "Explain the process in which employees submit ideas for improving production technology efficiency and the robot analyzes them. Use an emotion engine to reflect the employee's emotional state in the evaluation of the idea."

[0811] The present invention realizes highly accurate proposal evaluation by taking into consideration not only the content of the proposal but also the feelings of the submitter.

[0812] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0813] Step 1:

[0814] The device converts the user's input ideas into text data using voice recognition technology. Specifically, the device receives voice input through a built-in microphone, and the voice recognition engine converts this into text data. The input is voice data, and the output is text data.

[0815] Step 2:

[0816] The server receives the text data sent from the terminal. At this stage, the idea submitted by the user is saved in text format on the server. The input is text data, and the output is idea data stored in a database.

[0817] Step 3:

[0818] The server uses a natural language processing engine to preprocess the text data. Specifically, it performs stop word removal, stemming, and keyword extraction. The input is the text data, and the output is the preprocessed text data.

[0819] Step 4:

[0820] The server uses TF-IDF or Word2Vec to convert the preprocessed text data into a numerical vector, where the input is the preprocessed text data and the output is a numerical vector.

[0821] Step 5:

[0822] The server calculates the cosine similarity between the numeric vectors. Specifically, it calculates the similarity score using the inner product of each numeric vector and the vector norm. The input is the numeric vector, and the output is the similarity score.

[0823] Step 6:

[0824] The server extracts ideas whose similarity scores are below a certain threshold, where the input is the similarity score and the output is unique ideas.

[0825] Step 7:

[0826] The device uses a camera and microphone to analyze the user's facial expressions and voice. It uses an emotion engine to acquire the user's emotional data. The input is facial expression data and voice data, and the output is emotional data.

[0827] Step 8:

[0828] The server adjusts the evaluation of ideas based on the emotional data. Specifically, it increases the evaluation score of ideas with positive emotions. The inputs are the emotional data and unique idea data, and the output is the re-evaluated idea data.

[0829] Step 9:

[0830] The server notifies the factory manager of the final evaluation result of the unique idea. The input is the re-evaluated idea data, and the output is a notification message.

[0831] In this way, the system takes into account both the content of the proposal and the emotional state of the submitter, allowing it to efficiently extract and evaluate the most innovative ideas.

[0832] 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.

[0833] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0834] 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.

[0835] [Third embodiment]

[0836] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0837] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0838] 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).

[0839] 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.

[0840] 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.

[0841] 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).

[0842] 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.

[0843] 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.

[0844] 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.

[0845] 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.

[0846] 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.

[0847] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0848] This invention relates to a system that efficiently analyzes multiple proposals submitted in innovation contests and idea pitch events and automatically extracts innovative proposals. This system includes the following processing steps, and can consistently perform proposal submission, processing, and evaluation.

[0849] Program processing procedure

[0850] 1. Idea collection

[0851] Users submit ideas to innovation contests and idea pitch events via web forms or dedicated applications.

[0852] 2. Save your ideas in a database

[0853] The server receives the submitted ideas and stores them in a database. The stored data includes the idea text, submitter information, and submission date and time.

[0854] 3. Preprocessing ideas using natural language processing

[0855] The server retrieves unprocessed ideas from the database and performs the following natural language processing on the retrieved ideas.

[0856] Stop word removal: remove common meaningless words (e.g. "no", "wa", "ni", etc.).

[0857] Stemming: Extracting the base form of a word. (Example: "hashiru" → "hashiru")

[0858] Keyword extraction: Extract important words and phrases.

[0859] 4. Converting ideas into numerical vectors

[0860] The server converts the preprocessed idea text data into a numerical vector using the following techniques:

[0861] TF-IDF (Term Frequency-Inverse Document Frequency): Evaluates and quantifies the importance of each word.

[0862] Word2Vec: Generates numeric vectors based on word meanings.

[0863] 5. Similarity calculation in vector space

[0864] The server calculates the cosine similarity between the numeric vectors of all ideas, which allows it to calculate a similarity score between proposals and identify less relevant proposals.

[0865] 6. Extracting unique ideas

[0866] The server extracts ideas with similarity scores below a certain threshold. These are considered unique ideas that are significantly different from other proposals and are stored in a database as "unique ideas."

[0867] 7. Notices and Confirmations

[0868] The server notifies the judges and users, prompting them to review and evaluate the extracted unique ideas.

[0869] Specific examples

[0870] Example: Idea submission and processing

[0871] 1. A user submits two ideas using a web form, for example, "a new energy-efficient battery design" and "furniture made from recyclable materials."

[0872] 2. The server stores these ideas in a database and marks them as pending.

[0873] 3. The server performs natural language processing to remove stop words, stem, and extract key keywords from each idea.

[0874] Idea A Keywords: ["energy", "efficiency", "battery", "design"]

[0875] Idea B Keywords: ["recyclable", "material", "furniture"]

[0876] 4. The server generates a numeric vector based on the keyword.

[0877] Idea A's vector: [0.5, 0.7, 0.6, 0.8]

[0878] Idea B vector: [0.1, 0.3, 0.9, 0.4]

[0879] 5. The server calculates the cosine similarity and finds that the similarity score between these ideas is 0.2. Based on this low score, Idea A and Idea B are extracted as unique ideas.

[0880] 6. The server notifies the judges and prompts them to review and evaluate the extracted ideas.

[0881] This system makes it possible to efficiently extract innovative and unique ideas from a large number of proposals.

[0882] The processing flow will be explained below.

[0883] Step 1:

[0884] A user enters their idea in a web form or a dedicated application and clicks the submit button. The idea is entered in text format.

[0885] Step 2:

[0886] The terminal transmits the idea input by the user to the server, and the transmitted data includes the idea text, the user's identification information, the submission date and time, etc.

[0887] Step 3:

[0888] The server stores the received idea data in the database, and the stored data is marked with an unprocessed flag for the idea.

[0889] Step 4:

[0890] The server retrieves ideas with unprocessed flags from the database, and the retrieved ideas are subjected to natural language processing.

[0891] Step 5:

[0892] The server removes stop words from the idea text, which in the case of Japanese is a process to remove common words such as "no", "wa", and "ni".

[0893] Step 6:

[0894] The server performs stemming and converts words to their base form, for example, "run" and "ran" are converted to "run".

[0895] Step 7:

[0896] The server performs keyword extraction and lists important words and phrases, which are later converted into numerical vectors.

[0897] Step 8:

[0898] The server takes the preprocessed text and converts it into a numerical vector, using techniques such as TF-IDF or Word2Vec.

[0899] Step 9:

[0900] The server calculates the cosine similarity between the generated numeric vectors for all ideas. This similarity score evaluates the relative relevance of the proposals.

[0901] Step 10:

[0902] The server evaluates the similarity scores and extracts ideas with scores below a set threshold, which is used to identify ideas that are unique and different from other proposals.

[0903] Step 11:

[0904] The server stores the extracted unique ideas in a database and sets a special flag so that they are given priority access in subsequent evaluation processes.

[0905] Step 12:

[0906] The server sends notifications to users and judges, including a summary of the unique ideas extracted and a request for confirmation and evaluation.

[0907] Example 1

[0908] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0909] Conventional proposal extraction systems have had difficulty efficiently extracting innovative and unique proposals from a large number of proposals. Furthermore, the accuracy of proposal preprocessing and similarity calculations was insufficient, making it impossible to guarantee the quality of the extracted proposals. As a result, proposal evaluation took a long time, increasing the burden on judges and users.

[0910] 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.

[0911] In this invention, the server includes means for acquiring a plurality of submitted proposals, means for storing the acquired proposals in a database, means for acquiring unprocessed proposals from the database, means for preprocessing text data of the proposals using natural language processing, means for converting the preprocessed proposals into numerical vectors, means for calculating the similarity of each proposal in a vector space, means for extracting proposals whose similarity is equal to or less than a certain threshold based on the calculated similarity, and means for notifying the extracted proposals. This makes it possible to efficiently extract innovative and unique proposals from a large number of proposals.

[0912] "Proposal" means an idea or outline submitted by a User in an innovation contest or idea pitch event.

[0913] "Database" refers to a collection of information for storing and managing submitted proposals and related information (such as submitter information, submission date and time, etc.).

[0914] "Natural language processing" is a general term for technology that analyzes text data, understands its meaning, and extracts necessary information.

[0915] A "numeric vector" is a mathematical representation of text data or proposal content in a format that allows for calculations within a vector space.

[0916] "Cosine similarity" is a calculation method for evaluating the similarity between two numerical vectors, and uses the angle between the vectors.

[0917] "TF-IDF" stands for "Term Frequency-Inverse Document Frequency" and is a method for calculating and quantifying the importance of each word.

[0918] "Word2Vec" is a technology for representing the meaning of words as numerical vectors, and is a model that trains words that are semantically similar to each other so that they have similar vectors.

[0919] "Notification" refers to the action of sending information so that judges and users are aware of the proposal.

[0920] The present invention relates to a system that efficiently analyzes multiple proposals submitted in innovation contests and idea pitch events and automatically extracts innovative proposals. This system operates using the following hardware and software.

[0921] Hardware and software used

[0922] Server: A computer system for database management, natural language processing, numeric vector transformation, similarity calculation, and notification sending.

[0923] Database: An information management system for storing proposal data, user information, and similarity scores.

[0924] Natural language processing libraries: For example, use Python's nlp library to analyze text data.

[0925] Numerical vector transformation libraries: For example, use the scikit-learn library to perform TF-IDF and Word2Vec transformations.

[0926] Similarity calculation library: For example, use the scipy library to calculate cosine similarity.

[0927] Notification Sending API: API for sending emails and push notifications.

[0928] Example of a system

[0929] Example: Idea submission and processing

[0930] 1. A user submits an idea for a "new energy-efficient battery design" using a web form.

[0931] 2. The server receives the idea and stores it in the database in the format "Idea: New energy-efficient battery design, Submitted by: Username, Submitted date: 2023-10-01".

[0932] 3. The server retrieves raw ideas from the database and performs natural language processing, such as removing "no," "wa," and "ni" from the idea text, stemming "hashiru" to "so," and extracting keywords such as "energy," "efficiency," "battery," and "design."

[0933] 4. The server calculates the TF-IDF based on these keywords and generates a numerical vector, for example, "energy" is 0.7, "efficiency" is 0.8, "battery" is 0.9, and "design" is 1.0.

[0934] 5. The server calculates the cosine similarity between all ideas and obtains a similarity score (e.g., 0.2).

[0935] 6. The server extracts ideas with a similarity score below a threshold (e.g., 0.3) as unique ideas and stores them in the database. For example, it flags them as "unique ideas" using an SQL UPDATE statement.

[0936] 7. The server sends a notification email to the judges, prompting them to confirm and evaluate the unique ideas.

[0937] Example prompts for generative AI models

[0938] Here are some example prompts you can enter into your generative AI model:

[0939] "Please extract important keywords from the following text: 'New energy-efficient battery design'"

[0940] The present invention makes it possible to efficiently extract innovative and unique proposals from a large number of proposals, thereby improving the efficiency of the review and the quality of the evaluation.

[0941] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0942] Step 1:

[0943] A user submits an idea. Using a web form, they enter the title and body of the idea and click the "Submit" button. The input is the title and body of the idea, and the output is the idea sent to the server.

[0944] Step 2:

[0945] The server receives the submitted idea. This temporarily stores the submitted idea data (title, text, submitter information, submission date and time). The input is the idea data sent by the user, and the output is the data stored in the server's temporary memory.

[0946] Step 3:

[0947] The server saves the received idea data in the database. The server establishes a database connection and uses the SQL INSERT statement to store the idea data in persistent storage. As input, there is idea data stored in temporary memory, and as output, the idea data is saved in the database.

[0948] Step 4:

[0949] The server retrieves unprocessed ideas from the database. It uses a SQL SELECT statement to extract idea data that is marked as unprocessed. As input, it has idea data that is flagged as unprocessed, and as output, it extracts that data to the server.

[0950] Step 5:

[0951] The server performs natural language processing (NLP) to preprocess the idea data. It uses Python's NLP library to remove stop words, perform stemming, and extract keywords. The extracted idea data is input, and preprocessed idea data is generated as output. Specifically, it removes words such as "no," "wa," and "ni" from the main body of the idea, converts "hashiru" to "so," and extracts keywords such as "energy," "efficiency," "battery," and "design."

[0952] Step 6:

[0953] The server converts the preprocessed idea data into a numerical vector. Using Python's scikit-learn library, the conversion is performed using TF-IDF and Word2Vec. The preprocessed idea data is input, and a numerical vector is generated as output. Specifically, the importance of each keyword is evaluated and quantified.

[0954] Step 7:

[0955] The server calculates the cosine similarity between all idea vectors. The similarity calculation is performed using the scipy library. The input is a set of numerical vectors, and the output is a similarity score between each vector. Specifically, the angle between the vectors is calculated and the result is scored.

[0956] Step 8:

[0957] The server extracts unique ideas based on the similarity score. The server stores ideas with a similarity score below a certain threshold as "unique ideas" in a database. The input is the similarity score, and the output is data flagged as a unique idea.

[0958] Step 9:

[0959] The server sends notifications to judges and users. The server uses an email sending API to notify unique ideas that require evaluation. The input is the data of the unique idea, and the output is a notification email. Specifically, the body of the notification email contains the title of the idea and an evaluation link.

[0960] (Application example 1)

[0961] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0962] Proposing suggestions to improve work efficiency on-site requires a lot of time and effort. In particular, the process of analyzing information collected from the factory floor, extracting useful suggestions, and quickly communicating the results to workers is complex, which can delay responses when immediate action is required. In addition, traditional methods often include many overlapping or similar suggestions, which can result in useful suggestions being overlooked.

[0963] 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.

[0964] In this invention, the server includes a means for acquiring a plurality of submitted proposals, a means for converting the proposals into a numerical vector, a means for calculating similarity, a means for extracting proposals whose similarity is equal to or less than a certain threshold, and a means for using a smart device to present the extracted proposals. This makes it possible to quickly identify useful proposals for supporting improved work efficiency at the factory site and quickly notify workers.

[0965] The "means for acquiring a plurality of submitted proposals" refers to a means by which workers or engineers on the factory floor use smart devices to input proposals and transmit them to a server.

[0966] The "means for converting into a numerical vector" is a means for converting the proposed text data into a numerical vector using a specific algorithm (TF-IDF or Word2Vec).

[0967] The "means for calculating similarity" is a means for performing calculations to evaluate the similarity between proposals using the converted numerical vectors.

[0968] The "means for extracting proposals whose similarity is equal to or less than a certain threshold" is a means for identifying and extracting unique proposals that are significantly different from other proposals, based on the similarity scores between proposals.

[0969] The "means of using a smart device to present extracted suggestions" refers to a means of notifying and displaying the extracted unique suggestions to workers or engineers in real time through a smart device such as smart glasses.

[0970] The present invention is a proposed analysis system aimed at improving work efficiency at a factory site. A specific embodiment of this system is described below.

[0971] First, workers or engineers use smart glasses to voice-input improvements or new ideas they discover on-site. The voice-input content is converted into text data by the voice recognition software in the smart glasses. This text data is then sent to a server via a network. This process uses smart glasses (e.g., Google Glass) and their built-in voice recognition software (e.g., Google Speech-to-Text API).

[0972] The server stores the received suggestions in a database and then performs the following natural language processing: The server removes common meaningless words (stop words) from the suggested text and converts words to their root forms (stemming). This process is performed using the NLTK library. After that, the text data is converted into a numerical vector using the TF-IDF (Term Frequency-Inverse Document Frequency) or Word2Vec algorithm to extract important keywords. This is performed using the Scikit-learn library.

[0973] Next, the server calculates the cosine similarity between the vectors and calculates a similarity score for each proposal. It identifies proposals with a similarity score below a certain threshold (unique proposals) and extracts them. The Scikit-learn library is also used to calculate the similarity.

[0974] The extracted unique suggestions are sent to workers and engineers using smart glasses. The notification includes an overview of the suggestion and its evaluation, allowing workers to immediately check and evaluate the content of the suggestion. This enables quick responses on-site and is expected to lead to efficient work improvements.

[0975] As a specific use case, field worker A makes a suggestion by voice, "Design a new, energy-efficient battery." This suggestion is sent to the server, where it undergoes natural language processing and similarity calculations are performed. Similarly, worker B makes a suggestion, "Furniture made from recyclable materials." The server evaluates the similarity between these suggestions and identifies both as unique suggestions. This allows workers to instantly receive useful suggestions and incorporate them into their improvement activities.

[0976] An example of an input prompt for the generative AI model is as follows:

[0977] Next, enter the idea you want to collect. For example, enter a brief description of your idea, such as "new energy-efficient battery designs" or "furniture made from recyclable materials."

[0978] By using the above-described means, the present invention improves the efficiency of a series of processes from collection of proposals to analysis and notification at the factory site, thereby contributing to improved work efficiency.

[0979] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0980] Step 1:

[0981] The user uses the smart glasses to speak suggestions, which are converted into text by voice recognition software within the glasses, and then sent to the server.

[0982] Input: User's voice suggestions

[0983] Output: Proposals converted to text data

[0984] How it works: The smart glasses record the voice suggestions and convert them into text using voice recognition software, which then sends the text data over the network to a server.

[0985] Step 2:

[0986] The server stores the received proposal in a database. Data including details of the proposal (text, submitter information, submission date and time, etc.) is recorded in the database.

[0987] Input: Text data

[0988] Output: Proposal data stored in a database

[0989] What happens: The server receives the text data and stores it as a new record in the database. All necessary metadata is also saved at this stage.

[0990] Step 3:

[0991] The server retrieves the raw stored suggestions and performs natural language processing on them, specifically stop word removal, stemming, and keyword extraction.

[0992] Input: Proposal data retrieved from the database

[0993] Output: Preprocessed text data

[0994] Specific operation: The server uses the NLTK library to remove stop words and perform stemming, and performs keyword extraction using algorithms such as TF-IDF and Word2Vec.

[0995] Step 4:

[0996] The server converts the preprocessed text data of the proposals into a numerical vector using TF-IDF or Word2Vec.

[0997] Input: Preprocessed text data

[0998] Output: Numeric vector

[0999] Specific operation: The server converts text data into a numerical vector using TF-IDF or Word2Vec from the Scikit-learn library.

[1000] Step 5:

[1001] The server calculates the cosine similarity between all numeric vectors, which then produces a similarity score between the proposals.

[1002] Input: Numeric vector

[1003] Output: Similarity score

[1004] Specific operation: The server uses Scikit-learn's cosine_similarity method to calculate the cosine similarity between numeric vectors and obtain a similarity score.

[1005] Step 6:

[1006] The server extracts proposals with similarity scores below a certain threshold and identifies them as unique proposals.

[1007] Input: Similarity score

[1008] Output: A list of unique suggestions

[1009] Specific operation: The server evaluates the similarity score based on the set threshold and extracts a list of unique proposals that are significantly different from other proposals.

[1010] Step 7:

[1011] The server notifies the user of the extracted unique suggestions through the smart glasses, along with a summary of the suggestions and their ratings.

[1012] Input: A list of unique suggestions

[1013] Output: Notification of the suggestion displayed on the smart glasses

[1014] Specific operation: The server composes the extracted unique ideas into a notification message and sends it to the smart glasses via the network. The smart glasses then display the content of the suggestion to the user.

[1015] Through the above processing steps, the present invention makes it possible to quickly identify useful suggestions for improving work efficiency on the factory floor and notify workers of them.

[1016] 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.

[1017] The present invention achieves even more accurate evaluation by combining a system that analyzes multiple proposals submitted to innovation contests and idea pitch events and automatically extracts innovative proposals with an emotion engine that recognizes the user's emotions.

[1018] Program processing procedure

[1019] 1. Idea collection

[1020] Users submit ideas to innovation contests and idea pitch events via web forms or dedicated applications.

[1021] 2. Save your ideas in a database

[1022] The server receives the submitted ideas and stores them in a database. The stored data includes the idea text, submitter information, and submission date and time.

[1023] 3. Preprocessing ideas using natural language processing

[1024] The server retrieves unprocessed ideas from the database and performs the following natural language processing on the retrieved ideas.

[1025] Stop word removal: remove common meaningless words (e.g. "no", "wa", "ni", etc.).

[1026] Stemming: Extracting the base form of a word. (Example: "hashiru" → "hashiru")

[1027] Keyword extraction: Extract important words and phrases.

[1028] 4. Converting ideas into numerical vectors

[1029] The server converts the preprocessed idea text data into a numerical vector using the following techniques:

[1030] TF-IDF (Term Frequency-Inverse Document Frequency): Evaluates and quantifies the importance of each word.

[1031] Word2Vec: Generates numeric vectors based on word meanings.

[1032] 5. Similarity calculation in vector space

[1033] The server calculates the cosine similarity between the numeric vectors of all ideas, which allows it to calculate a similarity score between proposals and identify less relevant proposals.

[1034] 6. Extracting unique ideas

[1035] The server extracts ideas with similarity scores below a certain threshold. These are considered unique ideas that are significantly different from other proposals and are stored in a database as "unique ideas."

[1036] 7. Leveraging Emotional Engines

[1037] The server uses an emotion engine to acquire the user's emotional data, analyzing facial expressions and voice data when the user submits an idea, and quantifying the user's emotional state.

[1038] 8. Reflecting Emotional Data in Evaluation

[1039] The server influences the evaluation of the proposal based on the emotional data it acquires. For example, if the user has very positive emotions toward a proposal, it will increase the innovativeness rating of that proposal.

[1040] 9. Reevaluate unique ideas

[1041] The server performs re-evaluation using the emotion engine. Based on the emotion data, it recalculates the evaluation of unique ideas that have already been extracted and prioritizes them.

[1042] 10. Notices and Confirmations

[1043] The server sends notifications to users and judges, including a summary of the unique ideas extracted and a request for confirmation and evaluation.

[1044] Specific examples

[1045] Example: Idea submission and processing

[1046] 1. A user submits two ideas using a web form, for example, "a new energy-efficient battery design" and "furniture made from recyclable materials."

[1047] 2. The device sends these ideas to the server.

[1048] 3. The server stores these ideas in a database and marks them as pending.

[1049] 4. The server performs natural language processing to remove stop words, stem, and extract key keywords from each idea.

[1050] Idea A Keywords: ["energy", "efficiency", "battery", "design"]

[1051] Idea B Keywords: ["recyclable", "material", "furniture"]

[1052] 5. The server generates a numeric vector based on the keyword.

[1053] Idea A's vector: [0.5, 0.7, 0.6, 0.8]

[1054] Idea B vector: [0.1, 0.3, 0.9, 0.4]

[1055] 6. The server calculates the cosine similarity and finds that the similarity score between these ideas is 0.2. Based on this low score, Idea A and Idea B are extracted as unique ideas.

[1056] 7. The server uses an emotion engine to extract emotional data from facial expressions and voice when users submit ideas. For example, if a user is very excited, the idea will be rated higher.

[1057] 8. The server re-evaluates and prioritizes ideas based on sentiment data.

[1058] 9. The server notifies the judges and prompts them to review and evaluate the extracted ideas.

[1059] This system efficiently extracts innovative and unique ideas from a large number of proposals, and by evaluating them based on user emotional data, it is possible to achieve highly accurate evaluations.

[1060] The processing flow will be explained below.

[1061] Step 1:

[1062] A user enters their idea in a web form or a dedicated application and clicks the submit button. The idea is entered in text format.

[1063] Step 2:

[1064] The terminal transmits the idea input by the user to the server, and the transmitted data includes the idea text, the user's identification information, the submission date and time, etc.

[1065] Step 3:

[1066] The server stores the received idea data in the database, and the stored data is marked with an unprocessed flag for the idea.

[1067] Step 4:

[1068] The server retrieves ideas with unprocessed flags from the database, and the retrieved ideas are subjected to natural language processing.

[1069] Step 5:

[1070] The server removes stop words from the idea text, which in the case of Japanese is a process to remove common words such as "no", "wa", and "ni".

[1071] Step 6:

[1072] The server performs stemming and converts words to their base form, for example, "run" and "ran" are converted to "run".

[1073] Step 7:

[1074] The server performs keyword extraction and lists important words and phrases, which are later converted into numerical vectors.

[1075] Step 8:

[1076] The server takes the preprocessed text and converts it into a numerical vector, using techniques such as TF-IDF or Word2Vec.

[1077] Step 9:

[1078] The server calculates the cosine similarity between the generated numeric vectors for all ideas. This similarity score evaluates the relative relevance of the proposals.

[1079] Step 10:

[1080] The server evaluates the similarity scores and extracts ideas with scores below a set threshold, which is used to identify ideas that are unique and different from other proposals.

[1081] Step 11:

[1082] The server stores the extracted unique ideas in a database and sets a special flag so that they are given priority access in subsequent evaluation processes.

[1083] Step 12:

[1084] The server uses an emotion engine to acquire emotional data from the user's facial expressions and voice when submitting an idea. The emotional data is analyzed in real time and quantified to represent the user's emotional state (e.g., excitement, joy, surprise, etc.).

[1085] Step 13:

[1086] The server reflects the emotional data in the idea evaluation. Positive emotional data positively influences the evaluation score of the idea, while negative emotional data is used as an indicator to reconsider the evaluation.

[1087] Step 14:

[1088] The server uses the results of the emotion data to re-evaluate the unique ideas that have already been extracted. As a result of the re-evaluation, the priorities may change.

[1089] Step 15:

[1090] The server sends a notification to the user and judges. The notification includes a summary of the unique ideas extracted, as well as the results of a re-evaluation that takes into account the emotional data. The judges then make their final confirmation and evaluation based on this information.

[1091] The following is an explanation based on a specific example.

[1092] Example: Submitting and processing an idea

[1093] Step 1 and Step 2:

[1094] Users submit two ideas through a web form: "A new energy-efficient battery design" and "Furniture made from recyclable materials." The device then sends the ideas to the server.

[1095] Step 3 to Step 7:

[1096] The server stores these ideas in a database and performs natural language processing on each idea, removing stop words, stemming, and extracting important keywords.

[1097] For example, create a list of keywords such as "energy efficient battery design" and "recyclable materials furniture."

[1098] Steps 8 to 10:

[1099] The server converts keywords into numerical vectors and calculates the cosine similarity. Ideas with low similarity scores are extracted as unique ideas and stored in a database.

[1100] Step 11 to Step 14:

[1101] The server uses an emotion engine to obtain user emotion data and reflect it in the idea evaluation. For example, if a user is very excited when submitting a "new energy-efficient battery design," the evaluation score of the idea will be increased.

[1102] Re-evaluation is performed based on the emotional data, and the priority of the extracted unique ideas is recalculated.

[1103] Step 15:

[1104] The server notifies the judges and users and provides an overview of the unique ideas extracted and evaluation results that take into account the emotional data. The judges then make their final confirmation and evaluation based on this information.

[1105] This makes it possible to accurately evaluate and extract ideas that incorporate emotional data.

[1106] Example 2

[1107] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1108] Conventional idea evaluation systems have difficulty efficiently extracting innovative and unique proposals from a large number of proposals. Furthermore, they lack user emotional data, resulting in low evaluation accuracy. A system that can solve these issues and provide more accurate and efficient evaluations is needed.

[1109] 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.

[1110] In this invention, the server includes means for acquiring a plurality of submitted proposals, means for converting the proposals into numerical vectors, means for calculating the similarity of each proposal in a vector space, means for extracting proposals whose similarity is equal to or less than a certain threshold based on the calculated similarity, means for acquiring user emotion data using an emotion engine, and means for recalculating proposal evaluations based on the acquired emotion data. This makes it possible to efficiently extract innovative and unique proposals from a large number of proposals and to enable highly accurate evaluations that reflect the user's emotion data.

[1111] "Submitted Proposals" means ideas or proposals submitted by a User to the System through a contest or event.

[1112] "Means for obtaining" refers to the process or component by which the server receives and stores user input data.

[1113] "Means for converting to a numeric vector" refers to a process or algorithm for converting text data into a numeric vector format using natural language processing techniques.

[1114] "Means for calculating the similarity of each proposal in vector space" refers to a process or algorithm for calculating, for example, the cosine similarity between proposals converted into numerical vectors.

[1115] "Means for extracting proposals with similarity below a certain threshold" refers to a process or algorithm for selecting unique proposals that are significantly different from other proposals based on a calculated similarity score.

[1116] "Means for acquiring user emotional data using an emotion engine" refers to a process or component that uses technologies such as facial expression analysis and voice analysis to detect the user's emotional state and acquire it as data.

[1117] "Means for recalculating the evaluation of a proposal based on acquired emotional data" refers to a process or algorithm for correcting or recalculating the evaluation of the innovativeness or uniqueness of a proposal based on the emotional data.

[1118] This invention relates to a system that analyzes multiple proposals submitted to innovation contests and idea pitch events and automatically extracts innovative proposals. This system achieves more accurate evaluation by combining it with an emotion engine that recognizes the user's emotions.

[1119] The system uses the following hardware and software:

[1120] Server: Receives, stores, analyzes, and evaluates proposal data.

[1121] Terminal: Runs a web form or dedicated application that allows users to submit ideas.

[1122] Emotion engine: Facial expression and voice analysis technology to recognize user emotions (e.g., facial expression recognition software, voice recognition software).

[1123] System hardware and software configuration

[1124] 1. A means of retrieving multiple submitted proposals

[1125] Users submit ideas through a web form or a dedicated application, for example, by entering an idea such as "Design a new, energy-efficient battery."

[1126] 2. A means of converting proposals into numerical vectors

[1127] The server uses natural language processing technology to convert the text data of submitted ideas into numerical vectors, specifically using TF-IDF and Word2Vec to quantify the importance and meaning of words.

[1128] 3. A means to calculate the similarity of each proposal in the vector space

[1129] The server calculates the cosine similarity between the proposals converted into numerical vectors, which allows us to evaluate the relevance between the proposals.

[1130] 4. A method for extracting proposals whose similarity is below a certain threshold

[1131] The server extracts unique proposals that are significantly different from other proposals based on the calculated similarity scores, thus identifying innovative and unique proposals.

[1132] 5. Means of acquiring user emotion data using an emotion engine

[1133] The server uses an emotion engine to analyze facial expressions and voice data when a user submits a suggestion and quantify their emotional state. For example, a user who is very excited might be recorded as 80%.

[1134] 6. A method for recalculating the evaluation of proposals based on acquired emotional data

[1135] The server reflects the sentiment data and recalculates the evaluation of the proposal's innovativeness and uniqueness, so that the user's sentiment influences the proposal's final evaluation.

[1136] As a concrete example, consider a case where a user submits two ideas using a web form, say "designing a new energy-efficient battery" and "furniture made from recyclable materials." The device sends these ideas to a server, which stores them in a database and marks them as unprocessed. The server then performs natural language processing (NLP) to remove stop words from each idea, perform stemming, and extract key keywords. The server then generates a numerical vector based on the keywords and calculates the cosine similarity to calculate a similarity score between the proposals.

[1137] A particularly important step is to use an emotion engine to obtain user emotional data and re-evaluate proposals based on this, making it possible to select innovative proposals with greater accuracy than conventional systems.

[1138] Prompt Sentence Examples

[1139] A user uses a web form to submit a new energy-efficient battery design or furniture made from recyclable materials. How does the server process these and ultimately evaluate them?

[1140] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1141] Step 1:

[1142] Users submit ideas through a web form or a dedicated application. For example, they enter an idea such as "Design a new, energy-efficient battery." The entered idea data is sent from the device to a server.

[1143] Input: Idea data entered by the user

[1144] Output: Idea data sent from the device to the server

[1145] What happens: A user enters an idea into a web form, clicks the submit button, and the data is sent to the server.

[1146] Step 2:

[1147] The server stores the received idea data in a database, including the idea text, submitter information, and submission date and time.

[1148] Input: Idea data sent from the device

[1149] Output: Idea data stored in a database

[1150] Specific operation: The server receives the idea data and records it as a new entry in the database.

[1151] Step 3:

[1152] The server retrieves raw ideas from the database, and performs preprocessing on the text data of the ideas, such as removing stop words, stemming, and extracting keywords.

[1153] Input: Raw idea data retrieved from the database

[1154] Output: Preprocessed text data

[1155] Specific operation: For example, stop words are removed from "new energy-efficient battery design" to "new energy-efficient battery design" and converted to the basic form.

[1156] Step 4:

[1157] The server converts the preprocessed text data of ideas into numerical vectors, and quantifies the importance and meaning of words using TF-IDF and Word2Vec.

[1158] Input: Preprocessed text data

[1159] Output: Numerical vector data

[1160] Specific operation: For example, the sentence "New energy-efficient high-battery design" is converted into a numerical vector [0.5, 0.7, 0.6, 0.8].

[1161] Step 5:

[1162] The server calculates the cosine similarity between the numeric vectors of all ideas, which produces a similarity score between proposals and identifies less relevant proposals.

[1163] Input: Numeric vector data

[1164] Output: Similarity score

[1165] Specific operation: For example, the cosine similarity between the numeric vectors of idea A and idea B is calculated to be 0.2.

[1166] Step 6:

[1167] The server extracts ideas with similarity scores below a certain threshold and stores these ideas in a database as "unique ideas."

[1168] Input: Similarity score

[1169] Output: Unique Idea List

[1170] Specific behavior: Ideas with a similarity score of 0.3 or less are extracted and saved in the database with a new label.

[1171] Step 7:

[1172] The server uses an emotion engine to acquire the user's emotional data, analyzes facial expressions and voice data when the user submits an idea, and quantifies the user's emotional state.

[1173] Input: User's facial expression data, voice data

[1174] Output: Quantified emotion data

[1175] Specific operation: The user's facial expressions and voice recorded via the webcam and microphone are analyzed, and the "excitement" state is quantified as 80%.

[1176] Step 8:

[1177] Based on the sentiment data obtained by the server, a recalculation is performed to influence the rating of the proposals. Suggestions with high positive sentiment are given higher ratings and prioritized.

[1178] Input: Quantified emotion data, unique idea list

[1179] Output: Re-evaluated unique idea list

[1180] Specific operation: The evaluation scores are calculated again and it is determined that "Idea A is first place, and Idea B is second place."

[1181] Step 9:

[1182] The server sends notifications to users and judges, including a summary of the unique ideas extracted and a request for confirmation and evaluation.

[1183] Input: Reevaluated Unique Ideas List

[1184] Output: Notification message

[1185] Specific operation: The judges will be notified of the "New Energy-Efficient High-Battery Design" and will receive an email requesting their evaluation.

[1186] (Application example 2)

[1187] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1188] Conventional proposal evaluation systems primarily analyze the content of proposals to assess their innovativeness and uniqueness. However, they ignore human factors such as the submitter's emotional state and motivation at the time of proposal, and therefore are unable to fully assess whether the submitter is truly motivated and whether the idea is innovative. This creates a risk of overlooking promising ideas, and it is necessary to improve the accuracy of evaluation results.

[1189] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for acquiring multiple submitted proposals, means for converting the proposals into numerical vectors, means for calculating the similarity of each proposal in a vector space, means for extracting proposals whose similarity is equal to or less than a certain threshold based on the calculated similarity, and means for analyzing the submitter's emotions and adjusting the evaluation of the proposal based on the emotion data. This enables a comprehensive evaluation that takes into account the submitter's emotional state in addition to evaluating the content of the proposal itself.

[1190] A "proposal" is an idea or plan for a new technology or service submitted to an innovation contest or idea pitch event.

[1191] A "numeric vector" is a set of coordinate values ​​in a multidimensional space used to analyze the proposal.

[1192] A "vector space" is a multidimensional analytical domain for representing proposals as numerical vectors.

[1193] "Similarity" is an index that evaluates the relevance of proposal content based on the distance and angle between numerical vectors.

[1194] The "threshold" is a boundary value set for determining similarity, below which a proposal is extracted as unique.

[1195] "Emotion data" is information about the emotional state of the person submitting the proposal, obtained by analyzing the person's facial expression, voice, etc.

[1196] "Evaluation" is the process of determining the innovativeness and potential value of a proposal based on its content and sentiment data.

[1197] "Analysis" involves examining submitted proposals and sentiment data in detail to clarify their nature and relevance.

[1198] In this invention, a system centered on a server, a terminal, and a user operates in cooperation with each other. A specific embodiment of the system will be described below.

[1199] 1. Collect and save ideas

[1200] Users submit new ideas through a terminal that includes a specialized application and allows users to input each idea by voice, which is converted into text using speech recognition technology and sent to a server.

[1201] 2. Preprocessing using natural language processing

[1202] The server preprocesses the received ideas using a natural language processing engine (e.g., spaCy or scikit-learn), which includes stopword removal, stemming, and keyword extraction.

[1203] 3. Conversion to a numeric vector

[1204] The server converts the preprocessed text data into a numerical vector using TF-IDF or Word2Vec, which allows the characteristics of each idea to be expressed numerically.

[1205] 4. Calculating similarity and extracting unique ideas

[1206] The server calculates the cosine similarity in the vector space to evaluate the similarity of each idea. Ideas with similarity below a certain threshold are extracted as unique ideas.

[1207] 5. Acquisition and Analysis of Emotion Data

[1208] The server uses a device equipped with a camera and microphone to capture the submitter's emotional data. It uses an emotion engine (e.g., OpenCV and TensorFlow) to analyze facial expressions and tone of voice, allowing it to quantify the submitter's emotional state when submitting an idea.

[1209] 6. Reassessment based on emotional data

[1210] The server uses the emotion data to re-evaluate unique ideas and adjust their ratings: if excitement or enthusiasm is high, the idea's score will increase.

[1211] 7. Final Evaluation and Notification

[1212] The server extracts the most innovative and unique ideas and sends the results through a notification system to notify factory managers and other stakeholders.

[1213] Specific use cases

[1214] Suppose an employee submits an idea for a new material handling system. The idea is converted into text using a robotic voice recognition system and sent to a server. After natural language processing, the text is converted into a numerical vector using TF-IDF. The proposal is then evaluated by calculating cosine similarity and performing sentiment analysis.

[1215] Prompt Sentence Examples

[1216] "Explain the process in which employees submit ideas for improving production technology efficiency and the robot analyzes them. Use an emotion engine to reflect the employee's emotional state in the evaluation of the idea."

[1217] The present invention realizes highly accurate proposal evaluation by taking into consideration not only the content of the proposal but also the feelings of the submitter.

[1218] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1219] Step 1:

[1220] The device converts the user's input ideas into text data using voice recognition technology. Specifically, the device receives voice input through a built-in microphone, and the voice recognition engine converts this into text data. The input is voice data, and the output is text data.

[1221] Step 2:

[1222] The server receives the text data sent from the terminal. At this stage, the idea submitted by the user is saved in text format on the server. The input is text data, and the output is idea data stored in a database.

[1223] Step 3:

[1224] The server uses a natural language processing engine to preprocess the text data. Specifically, it performs stop word removal, stemming, and keyword extraction. The input is the text data, and the output is the preprocessed text data.

[1225] Step 4:

[1226] The server uses TF-IDF or Word2Vec to convert the preprocessed text data into a numerical vector, where the input is the preprocessed text data and the output is a numerical vector.

[1227] Step 5:

[1228] The server calculates the cosine similarity between the numeric vectors. Specifically, it calculates the similarity score using the inner product of each numeric vector and the vector norm. The input is the numeric vector, and the output is the similarity score.

[1229] Step 6:

[1230] The server extracts ideas whose similarity scores are below a certain threshold, where the input is the similarity score and the output is unique ideas.

[1231] Step 7:

[1232] The device uses a camera and microphone to analyze the user's facial expressions and voice. It uses an emotion engine to acquire the user's emotional data. The input is facial expression data and voice data, and the output is emotional data.

[1233] Step 8:

[1234] The server adjusts the evaluation of ideas based on the emotional data. Specifically, it increases the evaluation score of ideas with positive emotions. The inputs are the emotional data and unique idea data, and the output is the re-evaluated idea data.

[1235] Step 9:

[1236] The server notifies the factory manager of the final evaluation result of the unique idea. The input is the re-evaluated idea data, and the output is a notification message.

[1237] In this way, the system takes into account both the content of the proposal and the emotional state of the submitter, allowing it to efficiently extract and evaluate the most innovative ideas.

[1238] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1239] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1240] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1241] [Fourth embodiment]

[1242] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1243] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1244] 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).

[1245] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1246] 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.

[1247] 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).

[1248] 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.

[1249] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1250] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1251] 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.

[1252] 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.

[1253] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1254] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1255] This invention relates to a system that efficiently analyzes multiple proposals submitted in innovation contests and idea pitch events and automatically extracts innovative proposals. This system includes the following processing steps, and can consistently perform proposal submission, processing, and evaluation.

[1256] Program processing procedure

[1257] 1. Idea collection

[1258] Users submit ideas to innovation contests and idea pitch events via web forms or dedicated applications.

[1259] 2. Save your ideas in a database

[1260] The server receives the submitted ideas and stores them in a database. The stored data includes the idea text, submitter information, and submission date and time.

[1261] 3. Preprocessing ideas using natural language processing

[1262] The server retrieves unprocessed ideas from the database and performs the following natural language processing on the retrieved ideas.

[1263] Stop word removal: remove common meaningless words (e.g. "no", "wa", "ni", etc.).

[1264] Stemming: Extracting the base form of a word. (Example: "hashiru" → "hashiru")

[1265] Keyword extraction: Extract important words and phrases.

[1266] 4. Converting ideas into numerical vectors

[1267] The server converts the preprocessed idea text data into a numerical vector using the following techniques:

[1268] TF-IDF (Term Frequency-Inverse Document Frequency): Evaluates and quantifies the importance of each word.

[1269] Word2Vec: Generates numeric vectors based on word meanings.

[1270] 5. Similarity calculation in vector space

[1271] The server calculates the cosine similarity between the numeric vectors of all ideas, which allows it to calculate a similarity score between proposals and identify less relevant proposals.

[1272] 6. Extracting unique ideas

[1273] The server extracts ideas with similarity scores below a certain threshold. These are considered unique ideas that are significantly different from other proposals and are stored in a database as "unique ideas."

[1274] 7. Notices and Confirmations

[1275] The server notifies the judges and users, prompting them to review and evaluate the extracted unique ideas.

[1276] Specific examples

[1277] Example: Idea submission and processing

[1278] 1. A user submits two ideas using a web form, for example, "a new energy-efficient battery design" and "furniture made from recyclable materials."

[1279] 2. The server stores these ideas in a database and marks them as pending.

[1280] 3. The server performs natural language processing to remove stop words, stem, and extract key keywords from each idea.

[1281] Idea A Keywords: ["energy", "efficiency", "battery", "design"]

[1282] Idea B Keywords: ["recyclable", "material", "furniture"]

[1283] 4. The server generates a numeric vector based on the keyword.

[1284] Idea A's vector: [0.5, 0.7, 0.6, 0.8]

[1285] Idea B vector: [0.1, 0.3, 0.9, 0.4]

[1286] 5. The server calculates the cosine similarity and finds that the similarity score between these ideas is 0.2. Based on this low score, Idea A and Idea B are extracted as unique ideas.

[1287] 6. The server notifies the judges and prompts them to review and evaluate the extracted ideas.

[1288] This system makes it possible to efficiently extract innovative and unique ideas from a large number of proposals.

[1289] The processing flow will be explained below.

[1290] Step 1:

[1291] A user enters their idea in a web form or a dedicated application and clicks the submit button. The idea is entered in text format.

[1292] Step 2:

[1293] The terminal transmits the idea input by the user to the server, and the transmitted data includes the idea text, the user's identification information, the submission date and time, etc.

[1294] Step 3:

[1295] The server stores the received idea data in the database, and the stored data is marked with an unprocessed flag for the idea.

[1296] Step 4:

[1297] The server retrieves ideas with unprocessed flags from the database, and the retrieved ideas are subjected to natural language processing.

[1298] Step 5:

[1299] The server removes stop words from the idea text, which in the case of Japanese is a process to remove common words such as "no", "wa", and "ni".

[1300] Step 6:

[1301] The server performs stemming and converts words to their base form, for example, "run" and "ran" are converted to "run".

[1302] Step 7:

[1303] The server performs keyword extraction and lists important words and phrases, which are later converted into numerical vectors.

[1304] Step 8:

[1305] The server takes the preprocessed text and converts it into a numerical vector, using techniques such as TF-IDF or Word2Vec.

[1306] Step 9:

[1307] The server calculates the cosine similarity between the generated numeric vectors for all ideas. This similarity score evaluates the relative relevance of the proposals.

[1308] Step 10:

[1309] The server evaluates the similarity scores and extracts ideas with scores below a set threshold, which is used to identify ideas that are unique and different from other proposals.

[1310] Step 11:

[1311] The server stores the extracted unique ideas in a database and sets a special flag so that they are given priority access in subsequent evaluation processes.

[1312] Step 12:

[1313] The server sends notifications to users and judges, including a summary of the unique ideas extracted and a request for confirmation and evaluation.

[1314] Example 1

[1315] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1316] Conventional proposal extraction systems have had difficulty efficiently extracting innovative and unique proposals from a large number of proposals. Furthermore, the accuracy of proposal preprocessing and similarity calculations was insufficient, making it impossible to guarantee the quality of the extracted proposals. As a result, proposal evaluation took a long time, increasing the burden on judges and users.

[1317] 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.

[1318] In this invention, the server includes means for acquiring a plurality of submitted proposals, means for storing the acquired proposals in a database, means for acquiring unprocessed proposals from the database, means for preprocessing text data of the proposals using natural language processing, means for converting the preprocessed proposals into numerical vectors, means for calculating the similarity of each proposal in a vector space, means for extracting proposals whose similarity is equal to or less than a certain threshold based on the calculated similarity, and means for notifying the extracted proposals. This makes it possible to efficiently extract innovative and unique proposals from a large number of proposals.

[1319] "Proposal" means an idea or outline submitted by a User in an innovation contest or idea pitch event.

[1320] "Database" refers to a collection of information for storing and managing submitted proposals and related information (such as submitter information, submission date and time, etc.).

[1321] "Natural language processing" is a general term for technology that analyzes text data, understands its meaning, and extracts necessary information.

[1322] A "numeric vector" is a mathematical representation of text data or proposal content in a format that allows for calculations within a vector space.

[1323] "Cosine similarity" is a calculation method for evaluating the similarity between two numerical vectors, and uses the angle between the vectors.

[1324] "TF-IDF" stands for "Term Frequency-Inverse Document Frequency" and is a method for calculating and quantifying the importance of each word.

[1325] "Word2Vec" is a technology for representing the meaning of words as numerical vectors, and is a model that trains words that are semantically similar to each other so that they have similar vectors.

[1326] "Notification" refers to the action of sending information so that judges and users are aware of the proposal.

[1327] The present invention relates to a system that efficiently analyzes multiple proposals submitted in innovation contests and idea pitch events and automatically extracts innovative proposals. This system operates using the following hardware and software.

[1328] Hardware and software used

[1329] Server: A computer system for database management, natural language processing, numeric vector transformation, similarity calculation, and notification sending.

[1330] Database: An information management system for storing proposal data, user information, and similarity scores.

[1331] Natural language processing libraries: For example, use Python's nlp library to analyze text data.

[1332] Numerical vector transformation libraries: For example, use the scikit-learn library to perform TF-IDF and Word2Vec transformations.

[1333] Similarity calculation library: For example, use the scipy library to calculate cosine similarity.

[1334] Notification Sending API: API for sending emails and push notifications.

[1335] Example of a system

[1336] Example: Idea submission and processing

[1337] 1. A user submits an idea for a "new energy-efficient battery design" using a web form.

[1338] 2. The server receives the idea and stores it in the database in the format "Idea: New energy-efficient battery design, Submitted by: Username, Submitted date: 2023-10-01".

[1339] 3. The server retrieves raw ideas from the database and performs natural language processing, such as removing "no," "wa," and "ni" from the idea text, stemming "hashiru" to "so," and extracting keywords such as "energy," "efficiency," "battery," and "design."

[1340] 4. The server calculates the TF-IDF based on these keywords and generates a numerical vector, for example, "energy" is 0.7, "efficiency" is 0.8, "battery" is 0.9, and "design" is 1.0.

[1341] 5. The server calculates the cosine similarity between all ideas and obtains a similarity score (e.g., 0.2).

[1342] 6. The server extracts ideas with a similarity score below a threshold (e.g., 0.3) as unique ideas and stores them in the database. For example, it flags them as "unique ideas" using an SQL UPDATE statement.

[1343] 7. The server sends a notification email to the judges, prompting them to confirm and evaluate the unique ideas.

[1344] Example prompts for generative AI models

[1345] Here are some example prompts you can enter into your generative AI model:

[1346] "Please extract important keywords from the following text: 'New energy-efficient battery design'"

[1347] The present invention makes it possible to efficiently extract innovative and unique proposals from a large number of proposals, thereby improving the efficiency of the review and the quality of the evaluation.

[1348] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1349] Step 1:

[1350] A user submits an idea. Using a web form, they enter the title and body of the idea and click the "Submit" button. The input is the title and body of the idea, and the output is the idea sent to the server.

[1351] Step 2:

[1352] The server receives the submitted idea. This temporarily stores the submitted idea data (title, text, submitter information, submission date and time). The input is the idea data sent by the user, and the output is the data stored in the server's temporary memory.

[1353] Step 3:

[1354] The server saves the received idea data in the database. The server establishes a database connection and uses the SQL INSERT statement to store the idea data in persistent storage. As input, there is idea data stored in temporary memory, and as output, the idea data is saved in the database.

[1355] Step 4:

[1356] The server retrieves unprocessed ideas from the database. It uses a SQL SELECT statement to extract idea data that is marked as unprocessed. As input, it has idea data that is flagged as unprocessed, and as output, it extracts that data to the server.

[1357] Step 5:

[1358] The server performs natural language processing (NLP) to preprocess the idea data. It uses Python's NLP library to remove stop words, perform stemming, and extract keywords. The extracted idea data is input, and preprocessed idea data is generated as output. Specifically, it removes words such as "no," "wa," and "ni" from the main body of the idea, converts "hashiru" to "so," and extracts keywords such as "energy," "efficiency," "battery," and "design."

[1359] Step 6:

[1360] The server converts the preprocessed idea data into a numerical vector. Using Python's scikit-learn library, the conversion is performed using TF-IDF and Word2Vec. The preprocessed idea data is input, and a numerical vector is generated as output. Specifically, the importance of each keyword is evaluated and quantified.

[1361] Step 7:

[1362] The server calculates the cosine similarity between all idea vectors. The similarity calculation is performed using the scipy library. The input is a set of numerical vectors, and the output is a similarity score between each vector. Specifically, the angle between the vectors is calculated and the result is scored.

[1363] Step 8:

[1364] The server extracts unique ideas based on the similarity score. The server stores ideas with a similarity score below a certain threshold as "unique ideas" in a database. The input is the similarity score, and the output is data flagged as a unique idea.

[1365] Step 9:

[1366] The server sends notifications to judges and users. The server uses an email sending API to notify unique ideas that require evaluation. The input is the data of the unique idea, and the output is a notification email. Specifically, the body of the notification email contains the title of the idea and an evaluation link.

[1367] (Application example 1)

[1368] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1369] Proposing suggestions to improve work efficiency on-site requires a lot of time and effort. In particular, the process of analyzing information collected from the factory floor, extracting useful suggestions, and quickly communicating the results to workers is complex, which can delay responses when immediate action is required. In addition, traditional methods often include many overlapping or similar suggestions, which can result in useful suggestions being overlooked.

[1370] 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.

[1371] In this invention, the server includes a means for acquiring a plurality of submitted proposals, a means for converting the proposals into a numerical vector, a means for calculating similarity, a means for extracting proposals whose similarity is equal to or less than a certain threshold, and a means for using a smart device to present the extracted proposals. This makes it possible to quickly identify useful proposals for supporting improved work efficiency at the factory site and quickly notify workers.

[1372] The "means for acquiring a plurality of submitted proposals" refers to a means by which workers or engineers on the factory floor use smart devices to input proposals and transmit them to a server.

[1373] The "means for converting into a numerical vector" is a means for converting the proposed text data into a numerical vector using a specific algorithm (TF-IDF or Word2Vec).

[1374] The "means for calculating similarity" is a means for performing calculations to evaluate the similarity between proposals using the converted numerical vectors.

[1375] The "means for extracting proposals whose similarity is equal to or less than a certain threshold" is a means for identifying and extracting unique proposals that are significantly different from other proposals, based on the similarity scores between proposals.

[1376] The "means of using a smart device to present extracted suggestions" refers to a means of notifying and displaying the extracted unique suggestions to workers or engineers in real time through a smart device such as smart glasses.

[1377] The present invention is a proposed analysis system aimed at improving work efficiency at a factory site. A specific embodiment of this system is described below.

[1378] First, workers or engineers use smart glasses to voice-input improvements or new ideas they discover on-site. The voice-input content is converted into text data by the voice recognition software in the smart glasses. This text data is then sent to a server via a network. This process uses smart glasses (e.g., Google Glass) and their built-in voice recognition software (e.g., Google Speech-to-Text API).

[1379] The server stores the received suggestions in a database and then performs the following natural language processing: The server removes common meaningless words (stop words) from the suggested text and converts words to their root forms (stemming). This process is performed using the NLTK library. After that, the text data is converted into a numerical vector using the TF-IDF (Term Frequency-Inverse Document Frequency) or Word2Vec algorithm to extract important keywords. This is performed using the Scikit-learn library.

[1380] Next, the server calculates the cosine similarity between the vectors and calculates a similarity score for each proposal. It identifies proposals with a similarity score below a certain threshold (unique proposals) and extracts them. The Scikit-learn library is also used to calculate the similarity.

[1381] The extracted unique suggestions are sent to workers and engineers using smart glasses. The notification includes an overview of the suggestion and its evaluation, allowing workers to immediately check and evaluate the content of the suggestion. This enables quick responses on-site and is expected to lead to efficient work improvements.

[1382] As a specific use case, field worker A makes a suggestion by voice, "Design a new, energy-efficient battery." This suggestion is sent to the server, where it undergoes natural language processing and similarity calculations are performed. Similarly, worker B makes a suggestion, "Furniture made from recyclable materials." The server evaluates the similarity between these suggestions and identifies both as unique suggestions. This allows workers to instantly receive useful suggestions and incorporate them into their improvement activities.

[1383] An example of an input prompt for the generative AI model is as follows:

[1384] Next, enter the idea you want to collect. For example, enter a brief description of your idea, such as "new energy-efficient battery designs" or "furniture made from recyclable materials."

[1385] By using the above-described means, the present invention improves the efficiency of a series of processes from collection of proposals to analysis and notification at the factory site, thereby contributing to improved work efficiency.

[1386] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1387] Step 1:

[1388] The user uses the smart glasses to speak suggestions, which are converted into text by voice recognition software within the glasses, and then sent to the server.

[1389] Input: User's voice suggestions

[1390] Output: Proposals converted to text data

[1391] How it works: The smart glasses record the voice suggestions and convert them into text using voice recognition software, which then sends the text data over the network to a server.

[1392] Step 2:

[1393] The server stores the received proposal in a database. Data including details of the proposal (text, submitter information, submission date and time, etc.) is recorded in the database.

[1394] Input: Text data

[1395] Output: Proposal data stored in a database

[1396] What happens: The server receives the text data and stores it as a new record in the database. All necessary metadata is also saved at this stage.

[1397] Step 3:

[1398] The server retrieves the raw stored suggestions and performs natural language processing on them, specifically stop word removal, stemming, and keyword extraction.

[1399] Input: Proposal data retrieved from the database

[1400] Output: Preprocessed text data

[1401] Specific operation: The server uses the NLTK library to remove stop words and perform stemming, and performs keyword extraction using algorithms such as TF-IDF and Word2Vec.

[1402] Step 4:

[1403] The server converts the preprocessed text data of the proposals into a numerical vector using TF-IDF or Word2Vec.

[1404] Input: Preprocessed text data

[1405] Output: Numeric vector

[1406] Specific operation: The server converts text data into a numerical vector using TF-IDF or Word2Vec from the Scikit-learn library.

[1407] Step 5:

[1408] The server calculates the cosine similarity between all numeric vectors, which then produces a similarity score between the proposals.

[1409] Input: Numeric vector

[1410] Output: Similarity score

[1411] Specific operation: The server uses Scikit-learn's cosine_similarity method to calculate the cosine similarity between numeric vectors and obtain a similarity score.

[1412] Step 6:

[1413] The server extracts proposals with similarity scores below a certain threshold and identifies them as unique proposals.

[1414] Input: Similarity score

[1415] Output: A list of unique suggestions

[1416] Specific operation: The server evaluates the similarity score based on the set threshold and extracts a list of unique proposals that are significantly different from other proposals.

[1417] Step 7:

[1418] The server notifies the user of the extracted unique suggestions through the smart glasses, along with a summary of the suggestions and their ratings.

[1419] Input: A list of unique suggestions

[1420] Output: Notification of the suggestion displayed on the smart glasses

[1421] Specific operation: The server composes the extracted unique ideas into a notification message and sends it to the smart glasses via the network. The smart glasses then display the content of the suggestion to the user.

[1422] Through the above processing steps, the present invention makes it possible to quickly identify useful suggestions for improving work efficiency on the factory floor and notify workers of them.

[1423] 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.

[1424] The present invention achieves even more accurate evaluation by combining a system that analyzes multiple proposals submitted to innovation contests and idea pitch events and automatically extracts innovative proposals with an emotion engine that recognizes the user's emotions.

[1425] Program processing procedure

[1426] 1. Idea collection

[1427] Users submit ideas to innovation contests and idea pitch events via web forms or dedicated applications.

[1428] 2. Save your ideas in a database

[1429] The server receives the submitted ideas and stores them in a database. The stored data includes the idea text, submitter information, and submission date and time.

[1430] 3. Preprocessing ideas using natural language processing

[1431] The server retrieves unprocessed ideas from the database and performs the following natural language processing on the retrieved ideas.

[1432] Stop word removal: remove common meaningless words (e.g. "no", "wa", "ni", etc.).

[1433] Stemming: Extracting the base form of a word. (Example: "hashiru" → "hashiru")

[1434] Keyword extraction: Extract important words and phrases.

[1435] 4. Converting ideas into numerical vectors

[1436] The server converts the preprocessed idea text data into a numerical vector using the following techniques:

[1437] TF-IDF (Term Frequency-Inverse Document Frequency): Evaluates and quantifies the importance of each word.

[1438] Word2Vec: Generates numeric vectors based on word meanings.

[1439] 5. Similarity calculation in vector space

[1440] The server calculates the cosine similarity between the numeric vectors of all ideas, which allows it to calculate a similarity score between proposals and identify less relevant proposals.

[1441] 6. Extracting unique ideas

[1442] The server extracts ideas with similarity scores below a certain threshold. These are considered unique ideas that are significantly different from other proposals and are stored in a database as "unique ideas."

[1443] 7. Leveraging Emotional Engines

[1444] The server uses an emotion engine to acquire the user's emotional data, analyzing facial expressions and voice data when the user submits an idea, and quantifying the user's emotional state.

[1445] 8. Reflecting Emotional Data in Evaluation

[1446] The server influences the evaluation of the proposal based on the emotional data it acquires. For example, if the user has very positive emotions toward a proposal, it will increase the innovativeness rating of that proposal.

[1447] 9. Reevaluate unique ideas

[1448] The server performs re-evaluation using the emotion engine. Based on the emotion data, it recalculates the evaluation of unique ideas that have already been extracted and prioritizes them.

[1449] 10. Notices and Confirmations

[1450] The server sends notifications to users and judges, including a summary of the unique ideas extracted and a request for confirmation and evaluation.

[1451] Specific examples

[1452] Example: Idea submission and processing

[1453] 1. A user submits two ideas using a web form, for example, "a new energy-efficient battery design" and "furniture made from recyclable materials."

[1454] 2. The device sends these ideas to the server.

[1455] 3. The server stores these ideas in a database and marks them as pending.

[1456] 4. The server performs natural language processing to remove stop words, stem, and extract key keywords from each idea.

[1457] Idea A Keywords: ["energy", "efficiency", "battery", "design"]

[1458] Idea B Keywords: ["recyclable", "material", "furniture"]

[1459] 5. The server generates a numeric vector based on the keyword.

[1460] Idea A's vector: [0.5, 0.7, 0.6, 0.8]

[1461] Idea B vector: [0.1, 0.3, 0.9, 0.4]

[1462] 6. The server calculates the cosine similarity and finds that the similarity score between these ideas is 0.2. Based on this low score, Idea A and Idea B are extracted as unique ideas.

[1463] 7. The server uses an emotion engine to extract emotional data from facial expressions and voice when users submit ideas. For example, if a user is very excited, the idea will be rated higher.

[1464] 8. The server re-evaluates and prioritizes ideas based on sentiment data.

[1465] 9. The server notifies the judges and prompts them to review and evaluate the extracted ideas.

[1466] This system efficiently extracts innovative and unique ideas from a large number of proposals, and by evaluating them based on user emotional data, it is possible to achieve highly accurate evaluations.

[1467] The processing flow will be explained below.

[1468] Step 1:

[1469] A user enters their idea in a web form or a dedicated application and clicks the submit button. The idea is entered in text format.

[1470] Step 2:

[1471] The terminal transmits the idea input by the user to the server, and the transmitted data includes the idea text, the user's identification information, the submission date and time, etc.

[1472] Step 3:

[1473] The server stores the received idea data in the database, and the stored data is marked with an unprocessed flag for the idea.

[1474] Step 4:

[1475] The server retrieves ideas with unprocessed flags from the database, and the retrieved ideas are subjected to natural language processing.

[1476] Step 5:

[1477] The server removes stop words from the idea text, which in the case of Japanese is a process to remove common words such as "no", "wa", and "ni".

[1478] Step 6:

[1479] The server performs stemming and converts words to their base form, for example, "run" and "ran" are converted to "run".

[1480] Step 7:

[1481] The server performs keyword extraction and lists important words and phrases, which are later converted into numerical vectors.

[1482] Step 8:

[1483] The server takes the preprocessed text and converts it into a numerical vector, using techniques such as TF-IDF or Word2Vec.

[1484] Step 9:

[1485] The server calculates the cosine similarity between the generated numeric vectors for all ideas. This similarity score evaluates the relative relevance of the proposals.

[1486] Step 10:

[1487] The server evaluates the similarity scores and extracts ideas with scores below a set threshold, which is used to identify ideas that are unique and different from other proposals.

[1488] Step 11:

[1489] The server stores the extracted unique ideas in a database and sets a special flag so that they are given priority access in subsequent evaluation processes.

[1490] Step 12:

[1491] The server uses an emotion engine to acquire emotional data from the user's facial expressions and voice when submitting an idea. The emotional data is analyzed in real time and quantified to represent the user's emotional state (e.g., excitement, joy, surprise, etc.).

[1492] Step 13:

[1493] The server reflects the emotional data in the idea evaluation. Positive emotional data positively influences the evaluation score of the idea, while negative emotional data is used as an indicator to reconsider the evaluation.

[1494] Step 14:

[1495] The server uses the results of the emotion data to re-evaluate the unique ideas that have already been extracted. As a result of the re-evaluation, the priorities may change.

[1496] Step 15:

[1497] The server sends a notification to the user and judges. The notification includes a summary of the unique ideas extracted, as well as the results of a re-evaluation that takes into account the emotional data. The judges then make their final confirmation and evaluation based on this information.

[1498] The following is an explanation based on a specific example.

[1499] Example: Submitting and processing an idea

[1500] Step 1 and Step 2:

[1501] Users submit two ideas through a web form: "A new energy-efficient battery design" and "Furniture made from recyclable materials." The device then sends the ideas to the server.

[1502] Step 3 to Step 7:

[1503] The server stores these ideas in a database and performs natural language processing on each idea, removing stop words, stemming, and extracting important keywords.

[1504] For example, create a list of keywords such as "energy efficient battery design" and "recyclable materials furniture."

[1505] Steps 8 to 10:

[1506] The server converts keywords into numerical vectors and calculates the cosine similarity. Ideas with low similarity scores are extracted as unique ideas and stored in a database.

[1507] Step 11 to Step 14:

[1508] The server uses an emotion engine to obtain user emotion data and reflect it in the idea evaluation. For example, if a user is very excited when submitting a "new energy-efficient battery design," the evaluation score of the idea will be increased.

[1509] Re-evaluation is performed based on the emotional data, and the priority of the extracted unique ideas is recalculated.

[1510] Step 15:

[1511] The server notifies the judges and users and provides an overview of the unique ideas extracted and evaluation results that take into account the emotional data. The judges then make their final confirmation and evaluation based on this information.

[1512] This makes it possible to accurately evaluate and extract ideas that incorporate emotional data.

[1513] Example 2

[1514] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1515] Conventional idea evaluation systems have difficulty efficiently extracting innovative and unique proposals from a large number of proposals. Furthermore, they lack user emotional data, resulting in low evaluation accuracy. A system that can solve these issues and provide more accurate and efficient evaluations is needed.

[1516] 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.

[1517] In this invention, the server includes means for acquiring a plurality of submitted proposals, means for converting the proposals into numerical vectors, means for calculating the similarity of each proposal in a vector space, means for extracting proposals whose similarity is equal to or less than a certain threshold based on the calculated similarity, means for acquiring user emotion data using an emotion engine, and means for recalculating proposal evaluations based on the acquired emotion data. This makes it possible to efficiently extract innovative and unique proposals from a large number of proposals and to enable highly accurate evaluations that reflect the user's emotion data.

[1518] "Submitted Proposals" means ideas or proposals submitted by a User to the System through a contest or event.

[1519] "Means for obtaining" refers to the process or component by which the server receives and stores user input data.

[1520] "Means for converting to a numeric vector" refers to a process or algorithm for converting text data into a numeric vector format using natural language processing techniques.

[1521] "Means for calculating the similarity of each proposal in vector space" refers to a process or algorithm for calculating, for example, the cosine similarity between proposals converted into numerical vectors.

[1522] "Means for extracting proposals with similarity below a certain threshold" refers to a process or algorithm for selecting unique proposals that are significantly different from other proposals based on a calculated similarity score.

[1523] "Means for acquiring user emotional data using an emotion engine" refers to a process or component that uses technologies such as facial expression analysis and voice analysis to detect the user's emotional state and acquire it as data.

[1524] "Means for recalculating the evaluation of a proposal based on acquired emotional data" refers to a process or algorithm for correcting or recalculating the evaluation of the innovativeness or uniqueness of a proposal based on the emotional data.

[1525] This invention relates to a system that analyzes multiple proposals submitted to innovation contests and idea pitch events and automatically extracts innovative proposals. This system achieves more accurate evaluation by combining it with an emotion engine that recognizes the user's emotions.

[1526] The system uses the following hardware and software:

[1527] Server: Receives, stores, analyzes, and evaluates proposal data.

[1528] Terminal: Runs a web form or dedicated application that allows users to submit ideas.

[1529] Emotion engine: Facial expression and voice analysis technology to recognize user emotions (e.g., facial expression recognition software, voice recognition software).

[1530] System hardware and software configuration

[1531] 1. A means of retrieving multiple submitted proposals

[1532] Users submit ideas through a web form or a dedicated application, for example, by entering an idea such as "Design a new, energy-efficient battery."

[1533] 2. A means of converting proposals into numerical vectors

[1534] The server uses natural language processing technology to convert the text data of submitted ideas into numerical vectors, specifically using TF-IDF and Word2Vec to quantify the importance and meaning of words.

[1535] 3. A means to calculate the similarity of each proposal in the vector space

[1536] The server calculates the cosine similarity between the proposals converted into numerical vectors, which allows us to evaluate the relevance between the proposals.

[1537] 4. A method for extracting proposals whose similarity is below a certain threshold

[1538] The server extracts unique proposals that are significantly different from other proposals based on the calculated similarity scores, thus identifying innovative and unique proposals.

[1539] 5. Means of acquiring user emotion data using an emotion engine

[1540] The server uses an emotion engine to analyze facial expressions and voice data when a user submits a suggestion and quantify their emotional state. For example, a user who is very excited might be recorded as 80%.

[1541] 6. A method for recalculating the evaluation of proposals based on acquired emotional data

[1542] The server reflects the sentiment data and recalculates the evaluation of the proposal's innovativeness and uniqueness, so that the user's sentiment influences the proposal's final evaluation.

[1543] As a concrete example, consider a case where a user submits two ideas using a web form, say "designing a new energy-efficient battery" and "furniture made from recyclable materials." The device sends these ideas to a server, which stores them in a database and marks them as unprocessed. The server then performs natural language processing (NLP) to remove stop words from each idea, perform stemming, and extract key keywords. The server then generates a numerical vector based on the keywords and calculates the cosine similarity to calculate a similarity score between the proposals.

[1544] A particularly important step is to use an emotion engine to obtain user emotional data and re-evaluate proposals based on this, making it possible to select innovative proposals with greater accuracy than conventional systems.

[1545] Prompt Sentence Examples

[1546] A user uses a web form to submit a new energy-efficient battery design or furniture made from recyclable materials. How does the server process these and ultimately evaluate them?

[1547] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1548] Step 1:

[1549] Users submit ideas through a web form or a dedicated application. For example, they enter an idea such as "Design a new, energy-efficient battery." The entered idea data is sent from the device to a server.

[1550] Input: Idea data entered by the user

[1551] Output: Idea data sent from the device to the server

[1552] What happens: A user enters an idea into a web form, clicks the submit button, and the data is sent to the server.

[1553] Step 2:

[1554] The server stores the received idea data in a database, including the idea text, submitter information, and submission date and time.

[1555] Input: Idea data sent from the device

[1556] Output: Idea data stored in a database

[1557] Specific operation: The server receives the idea data and records it as a new entry in the database.

[1558] Step 3:

[1559] The server retrieves raw ideas from the database, and performs preprocessing on the text data of the ideas, such as removing stop words, stemming, and extracting keywords.

[1560] Input: Raw idea data retrieved from the database

[1561] Output: Preprocessed text data

[1562] Specific operation: For example, stop words are removed from "new energy-efficient battery design" to "new energy-efficient battery design" and converted to the basic form.

[1563] Step 4:

[1564] The server converts the preprocessed text data of ideas into numerical vectors, and quantifies the importance and meaning of words using TF-IDF and Word2Vec.

[1565] Input: Preprocessed text data

[1566] Output: Numerical vector data

[1567] Specific operation: For example, the sentence "New energy-efficient high-battery design" is converted into a numerical vector [0.5, 0.7, 0.6, 0.8].

[1568] Step 5:

[1569] The server calculates the cosine similarity between the numeric vectors of all ideas, which produces a similarity score between proposals and identifies less relevant proposals.

[1570] Input: Numeric vector data

[1571] Output: Similarity score

[1572] Specific operation: For example, the cosine similarity between the numeric vectors of idea A and idea B is calculated to be 0.2.

[1573] Step 6:

[1574] The server extracts ideas with similarity scores below a certain threshold and stores these ideas in a database as "unique ideas."

[1575] Input: Similarity score

[1576] Output: Unique Idea List

[1577] Specific behavior: Ideas with a similarity score of 0.3 or less are extracted and saved in the database with a new label.

[1578] Step 7:

[1579] The server uses an emotion engine to acquire the user's emotional data, analyzes facial expressions and voice data when the user submits an idea, and quantifies the user's emotional state.

[1580] Input: User's facial expression data, voice data

[1581] Output: Quantified emotion data

[1582] Specific operation: The user's facial expressions and voice recorded via the webcam and microphone are analyzed, and the "excitement" state is quantified as 80%.

[1583] Step 8:

[1584] Based on the sentiment data obtained by the server, a recalculation is performed to influence the rating of the proposals. Suggestions with high positive sentiment are given higher ratings and prioritized.

[1585] Input: Quantified emotion data, unique idea list

[1586] Output: Re-evaluated unique idea list

[1587] Specific operation: The evaluation scores are calculated again and it is determined that "Idea A is first place, and Idea B is second place."

[1588] Step 9:

[1589] The server sends notifications to users and judges, including a summary of the unique ideas extracted and a request for confirmation and evaluation.

[1590] Input: Reevaluated Unique Ideas List

[1591] Output: Notification message

[1592] Specific operation: The judges will be notified of the "New Energy-Efficient High-Battery Design" and will receive an email requesting their evaluation.

[1593] (Application example 2)

[1594] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1595] Conventional proposal evaluation systems primarily analyze the content of proposals to assess their innovativeness and uniqueness. However, they ignore human factors such as the submitter's emotional state and motivation at the time of proposal, and therefore are unable to fully assess whether the submitter is truly motivated and whether the idea is innovative. This creates a risk of overlooking promising ideas, and it is necessary to improve the accuracy of evaluation results.

[1596] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for acquiring multiple submitted proposals, means for converting the proposals into numerical vectors, means for calculating the similarity of each proposal in a vector space, means for extracting proposals whose similarity is equal to or less than a certain threshold based on the calculated similarity, and means for analyzing the submitter's emotions and adjusting the evaluation of the proposal based on the emotion data. This enables a comprehensive evaluation that takes into account the submitter's emotional state in addition to evaluating the content of the proposal itself.

[1597] A "proposal" is an idea or plan for a new technology or service submitted to an innovation contest or idea pitch event.

[1598] A "numeric vector" is a set of coordinate values ​​in a multidimensional space used to analyze the proposal.

[1599] A "vector space" is a multidimensional analytical domain for representing proposals as numerical vectors.

[1600] "Similarity" is an index that evaluates the relevance of proposal content based on the distance and angle between numerical vectors.

[1601] The "threshold" is a boundary value set for determining similarity, below which a proposal is extracted as unique.

[1602] "Emotion data" is information about the emotional state of the person submitting the proposal, obtained by analyzing the person's facial expression, voice, etc.

[1603] "Evaluation" is the process of determining the innovativeness and potential value of a proposal based on its content and sentiment data.

[1604] "Analysis" involves examining submitted proposals and sentiment data in detail to clarify their nature and relevance.

[1605] In this invention, a system centered on a server, a terminal, and a user operates in cooperation with each other. A specific embodiment of the system will be described below.

[1606] 1. Collect and save ideas

[1607] Users submit new ideas through a terminal that includes a specialized application and allows users to input each idea by voice, which is converted into text using speech recognition technology and sent to a server.

[1608] 2. Preprocessing using natural language processing

[1609] The server preprocesses the received ideas using a natural language processing engine (e.g., spaCy or scikit-learn), which includes stopword removal, stemming, and keyword extraction.

[1610] 3. Conversion to a numeric vector

[1611] The server converts the preprocessed text data into a numerical vector using TF-IDF or Word2Vec, which allows the characteristics of each idea to be expressed numerically.

[1612] 4. Calculating similarity and extracting unique ideas

[1613] The server calculates the cosine similarity in the vector space to evaluate the similarity of each idea. Ideas with similarity below a certain threshold are extracted as unique ideas.

[1614] 5. Acquisition and Analysis of Emotion Data

[1615] The server uses a device equipped with a camera and microphone to capture the submitter's emotional data. It uses an emotion engine (e.g., OpenCV and TensorFlow) to analyze facial expressions and tone of voice, allowing it to quantify the submitter's emotional state when submitting an idea.

[1616] 6. Reassessment based on emotional data

[1617] The server uses the emotion data to re-evaluate unique ideas and adjust their ratings: if excitement or enthusiasm is high, the idea's score will increase.

[1618] 7. Final Evaluation and Notification

[1619] The server extracts the most innovative and unique ideas and sends the results through a notification system to notify factory managers and other stakeholders.

[1620] Specific use cases

[1621] Suppose an employee submits an idea for a new material handling system. The idea is converted into text using a robotic voice recognition system and sent to a server. After natural language processing, the text is converted into a numerical vector using TF-IDF. The proposal is then evaluated by calculating cosine similarity and performing sentiment analysis.

[1622] Prompt Sentence Examples

[1623] "Explain the process in which employees submit ideas for improving production technology efficiency and the robot analyzes them. Use an emotion engine to reflect the employee's emotional state in the evaluation of the idea."

[1624] The present invention realizes highly accurate proposal evaluation by taking into consideration not only the content of the proposal but also the feelings of the submitter.

[1625] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1626] Step 1:

[1627] The device converts the user's input ideas into text data using voice recognition technology. Specifically, the device receives voice input through a built-in microphone, and the voice recognition engine converts this into text data. The input is voice data, and the output is text data.

[1628] Step 2:

[1629] The server receives the text data sent from the terminal. At this stage, the idea submitted by the user is saved in text format on the server. The input is text data, and the output is idea data stored in a database.

[1630] Step 3:

[1631] The server uses a natural language processing engine to preprocess the text data. Specifically, it performs stop word removal, stemming, and keyword extraction. The input is the text data, and the output is the preprocessed text data.

[1632] Step 4:

[1633] The server uses TF-IDF or Word2Vec to convert the preprocessed text data into a numerical vector, where the input is the preprocessed text data and the output is a numerical vector.

[1634] Step 5:

[1635] The server calculates the cosine similarity between the numeric vectors. Specifically, it calculates the similarity score using the inner product of each numeric vector and the vector norm. The input is the numeric vector, and the output is the similarity score.

[1636] Step 6:

[1637] The server extracts ideas whose similarity scores are below a certain threshold, where the input is the similarity score and the output is unique ideas.

[1638] Step 7:

[1639] The device uses a camera and microphone to analyze the user's facial expressions and voice. It uses an emotion engine to acquire the user's emotional data. The input is facial expression data and voice data, and the output is emotional data.

[1640] Step 8:

[1641] The server adjusts the evaluation of ideas based on the emotional data. Specifically, it increases the evaluation score of ideas with positive emotions. The inputs are the emotional data and unique idea data, and the output is the re-evaluated idea data.

[1642] Step 9:

[1643] The server notifies the factory manager of the final evaluation result of the unique idea. The input is the re-evaluated idea data, and the output is a notification message.

[1644] In this way, the system takes into account both the content of the proposal and the emotional state of the submitter, allowing it to efficiently extract and evaluate the most innovative ideas.

[1645] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1646] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1647] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1648] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1649] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1650] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1651] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1652] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1653] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1654] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1655] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1656] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1657] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1658] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1659] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1660] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1661] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1662] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1663] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1664] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1665] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1666] The following is further disclosed regarding the above embodiment.

[1667] (Claim 1)

[1668] a means for obtaining the plurality of submitted proposals;

[1669] a means of converting those proposals into numerical vectors;

[1670] a means for calculating the similarity of each proposal in the vector space;

[1671] A means for extracting proposals whose similarity is equal to or less than a certain threshold based on the calculated similarity;

[1672] A system including:

[1673] (Claim 2)

[1674] 2. The system according to claim 1, wherein cosine similarity is used to calculate the similarity.

[1675] (Claim 3)

[1676] The system of claim 1, wherein TF-IDF or Word2Vec is used for conversion to a numerical vector.

[1677] "Example 1"

[1678] (Claim 1)

[1679] a means for obtaining the plurality of submitted proposals;

[1680] a means for storing the obtained suggestions in a database;

[1681] a means for retrieving raw suggestions from the database; and

[1682] a means for preprocessing the proposal text data using natural language processing;

[1683] a means for converting the preprocessed proposals into a numeric vector;

[1684] a means for calculating the similarity of each proposal in the vector space;

[1685] A means for extracting proposals whose similarity is equal to or less than a certain threshold based on the calculated similarity;

[1686] a means for notifying the extracted suggestions;

[1687] A system including:

[1688] (Claim 2)

[1689] 2. The system according to claim 1, wherein cosine similarity is used to calculate the similarity.

[1690] (Claim 3)

[1691] The system of claim 1, wherein TF-IDF or Word2Vec is used for conversion to a numerical vector.

[1692] "Application Example 1"

[1693] (Claim 1)

[1694] a means for obtaining the plurality of submitted proposals;

[1695] a means of converting those proposals into numerical vectors;

[1696] a means for calculating the similarity of each proposal in the vector space;

[1697] A means for extracting proposals whose similarity is equal to or less than a certain threshold based on the calculated similarity;

[1698] a means for using a smart device to present the extracted suggestions;

[1699] A system including:

[1700] (Claim 2)

[1701] 2. The system according to claim 1, wherein cosine similarity is used to calculate the similarity.

[1702] (Claim 3)

[1703] The system of claim 1, wherein TF-IDF or Word2Vec is used for conversion to a numerical vector.

[1704] "Example 2: Combining Emotion Engines"

[1705] (Claim 1)

[1706] a means for obtaining the plurality of submitted proposals;

[1707] a means of converting those proposals into numerical vectors;

[1708] a means for calculating the similarity of each proposal in the vector space;

[1709] A means for extracting proposals whose similarity is equal to or less than a certain threshold based on the calculated similarity;

[1710] means for acquiring user emotion data using an emotion engine;

[1711] a means for recalculating the evaluation of the proposal based on the acquired emotion data;

[1712] A system including:

[1713] (Claim 2)

[1714] 2. The system according to claim 1, wherein cosine similarity is used to calculate the similarity.

[1715] (Claim 3)

[1716] The system of claim 1, wherein TF-IDF or Word2Vec is used for conversion to a numerical vector.

[1717] "Application example 2 when combining emotion engines"

[1718] (Claim 1)

[1719] a means for obtaining the plurality of submitted proposals;

[1720] a means of converting those proposals into numerical vectors;

[1721] a means for calculating the similarity of each proposal in the vector space;

[1722] A means for extracting proposals whose similarity is equal to or less than a certain threshold based on the calculated similarity;

[1723] a means for analyzing the submitter's sentiment and adjusting the evaluation of the proposal based on the sentiment data;

[1724] A system including:

[1725] (Claim 2)

[1726] 2. The system according to claim 1, wherein cosine similarity is used to calculate the similarity.

[1727] (Claim 3)

[1728] The system of claim 1, wherein TF-IDF or Word2Vec is used for conversion to a numerical vector. [Explanation of symbols]

[1729] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. a means for obtaining the plurality of submitted proposals; a means of converting those proposals into numerical vectors; a means for calculating the similarity of each proposal in the vector space; A means for extracting proposals whose similarity is equal to or less than a certain threshold based on the calculated similarity; A system including:

2. The system according to claim 1, wherein the cosine similarity is used to calculate the similarity.

3. The system according to claim 1, wherein TF-IDF or Word2Vec is used for conversion to a numerical vector.

Images