WO2023119460A1 - Système de prédiction de question, procédé de prédiction de question et support d'enregistrement de programme - Google Patents

Système de prédiction de question, procédé de prédiction de question et support d'enregistrement de programme Download PDF

Info

Publication number
WO2023119460A1
WO2023119460A1 PCT/JP2021/047466 JP2021047466W WO2023119460A1 WO 2023119460 A1 WO2023119460 A1 WO 2023119460A1 JP 2021047466 W JP2021047466 W JP 2021047466W WO 2023119460 A1 WO2023119460 A1 WO 2023119460A1
Authority
WO
WIPO (PCT)
Prior art keywords
shareholders
question
general meeting
prediction
questions
Prior art date
Application number
PCT/JP2021/047466
Other languages
English (en)
Japanese (ja)
Inventor
春華 茂刈
猛 伊藤
Original Assignee
日本電気株式会社
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 日本電気株式会社 filed Critical 日本電気株式会社
Priority to PCT/JP2021/047466 priority Critical patent/WO2023119460A1/fr
Publication of WO2023119460A1 publication Critical patent/WO2023119460A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"

Definitions

  • the present invention relates to a question prediction system and the like.
  • the assumed question-and-answer search system of Patent Document 1 receives an input of search conditions, it identifies a question corresponding to the search condition from the assumed question-and-answer database, and outputs the question content and the corresponding answer content.
  • the purpose is to provide a question prediction system etc. that can predict questions according to the situation of the company holding the general meeting of shareholders.
  • the question prediction system of the present invention includes acquisition means for acquiring information related to a company holding a general meeting of shareholders as general meeting of shareholders related information, and predicting questions at the general meeting of shareholders from the information related to the general meeting of shareholders. Prediction means for predicting questions at the general meeting of shareholders using a question prediction model based on the information related to the general meeting of shareholders acquired by the acquisition means, and output means for outputting the results of the prediction.
  • the question prediction method of the present invention obtains information related to a company that holds a general meeting of shareholders as general meeting of shareholders related information, and uses a question prediction model that predicts questions at the general meeting of shareholders from the information related to the general meeting of shareholders. Based on the related information, it predicts the questions at the shareholders' meeting and outputs the results of the prediction.
  • the program recording medium of the present invention obtains information related to a company that holds a general meeting of shareholders by using a process of acquiring information related to the general meeting of shareholders as general meeting of shareholders related information and a question prediction model that predicts questions at the general meeting of shareholders based on the information related to the general meeting of shareholders.
  • a question prediction program is recorded that causes a computer to execute a process of predicting questions at a general meeting of shareholders and a process of outputting the results of the prediction based on information related to the general meeting of shareholders.
  • questions can be predicted according to the situation of the company holding the general meeting of shareholders.
  • FIG. 1 is a diagram showing an overview of the configuration of a first embodiment of the present invention
  • FIG. BRIEF DESCRIPTION OF THE DRAWINGS It is a figure which shows the example of a structure of the question prediction system of the 1st Embodiment of this invention. It is a figure which shows the example of the display screen of the prediction result of the question of the 1st Embodiment of this invention. It is a figure which shows the example of the display screen of the prediction result of the question of the 1st Embodiment of this invention. It is a figure which shows the example of the operation
  • FIG. 1 is a diagram showing an overview of the configuration of an information processing system according to this embodiment.
  • the information processing system of this embodiment includes, as an example, a question prediction system 10 and a terminal device 20 .
  • the question prediction system 10 and the terminal device 20 are connected via a network. Also, a plurality of terminal devices 20 may be provided.
  • the question prediction system 10 is a system that uses a question prediction model to predict questions that will be asked at shareholders' meetings.
  • the question prediction model is a trained model that receives information related to the general meeting of shareholders as input and outputs questions expected at the general meeting of shareholders as prediction results.
  • the question prediction system 10 may predict questions using a plurality of question prediction models according to the company where the shareholders' meeting is held and the contents of the questions.
  • a question prediction model is generated by learning the relationship between shareholder meeting-related information and questions at the shareholder meeting.
  • the question prediction system 10 may predict questions using a question prediction model already generated in an external system.
  • Shareholder meeting-related information is information related to companies that hold shareholder meetings.
  • Information related to companies holding general shareholder meetings includes, for example, financial information, stock prices, valuations of corporate bonds, news, external announcements, postings about companies, advertisements, articles, analyst ratings, think tanks, etc. This is information on one or more items of information on analysis results, industry trends, trends of other companies, and corporate reputation.
  • Information related to a company holding a shareholders' meeting may include materials published by government offices that affect the company's business performance. Information related to a company holding a shareholders' meeting is not limited to the above examples.
  • External announcements include, for example, press releases, posting on the website, sending emails, transmitting information to SNS (Social Networking Services), conference presentations, briefings for investors, briefings for mass media, and other acts of transmitting information outside the company.
  • the external announcement is not limited to the above example.
  • a company's reputation is, for example, information posted on a website, information posted on an SNS, and questionnaire results.
  • Corporate reputation is not limited to the above examples.
  • the terminal device 20 is used, for example, as an operational terminal device for inputting data necessary for predicting questions and viewing results when predicting questions to be asked at a shareholders' meeting.
  • a person in charge of preparing for the general meeting of shareholders in the legal department of a company that holds the general meeting of shareholders accesses the question prediction system 10 via the network by operating the terminal device 20, and predicts questions at the general meeting of shareholders. conduct.
  • a plurality of persons in charge may access the question prediction system 10 by operating their respective terminal devices 20 .
  • the person who operates the terminal device 20 is not limited to the person in charge of the company's legal department.
  • FIG. 2 is a diagram showing an example of the configuration of the question prediction system 10.
  • the question prediction system 10 includes an acquisition unit 11 , a prediction unit 12 , a prediction model generation unit 13 , an output unit 14 and a storage unit 15 .
  • the acquisition unit 11 acquires information related to companies that hold shareholders' meetings as shareholders' meeting-related information.
  • the acquisition unit 11 acquires, for example, information related to the general meeting of shareholders during a period that is likely to be reflected in questions.
  • the period that is likely to be reflected in the question is the period during which the information related to the general meeting of shareholders was updated or announced, and is likely to be referred to by the person asking the question.
  • the period that is likely to be reflected in the question is set by, for example, the update frequency of information.
  • the acquiring unit 11 selects information related to the general meeting of shareholders for a period that is likely to be reflected in questions, such as information updated after the previous general meeting of shareholders. or obtain publicly available shareholder meeting-related information.
  • the general meeting of shareholders related information is information that is frequently updated or published, such as information on news or SNS
  • the acquisition unit 11 selects, for example, the holding of the general meeting of shareholders as the information related to the general meeting of shareholders during the period that is likely to be reflected in the question. Obtain information that has been updated or published since 1 month before the date.
  • the period that is likely to be reflected in the question is not limited to the above example.
  • the acquisition unit 11 may acquire information related to the general meeting of shareholders from the point in time prior to the time of the general meeting of shareholders to the time of the general meeting of shareholders.
  • the obtaining unit 11 obtains, from the terminal device 20, information related to the general meeting of shareholders input to the terminal device 20 by the operator's operation, for example.
  • the acquisition unit 11 may acquire the general meeting of shareholders related information from an information server connected via a network.
  • the acquisition unit 11 stores the acquired information related to the general meeting of shareholders in the storage unit 15 .
  • the acquisition unit 11 may acquire the general meeting of shareholders related information and question data.
  • the acquisition unit 11 uses, for example, data of questions at general shareholders' meetings held in the past and shareholder information during the period at which the questions were held at the general shareholders' meeting where the questions were likely to be reflected in the questions. Get general meeting related information.
  • the acquisition unit 11 acquires the general meeting of shareholders-related information and question data according to the prediction target. For example, when a question prediction model is generated for each company attribute, information related to shareholders' meetings and question data for companies with matching attributes are acquired. Companies with matching attributes may include companies with similar attributes.
  • the acquisition unit 11 acquires from the terminal device 20, for example, the information related to the general meeting of shareholders and the question data input to the terminal device 20 by the operator's operation.
  • the acquiring unit 11 may acquire the general meeting of shareholders related information and question data from an information server connected via a network.
  • the acquiring unit 11 stores the acquired information related to the shareholders' meeting and the data of the question, which are used for generating the question prediction model, in the storage unit 15 .
  • the acquisition unit 11 acquires the selection result of the question prediction model when a different question prediction model is used for each target of prediction when the question is predicted. For example, the acquisition unit 11 acquires from the terminal device 20 the selection of the question prediction model input to the terminal device 20 by the operator's operation.
  • the prediction unit 12 predicts questions at the general meeting of shareholders based on the information related to the general meeting of shareholders acquired by the acquiring unit 11, using a question prediction model that has learned the relationship between information related to the general meeting of shareholders and questions.
  • the prediction unit 12 uses, for example, a question prediction model generated by the prediction model generation unit 13 to predict questions at the shareholders meeting.
  • the prediction unit 12 may use a question prediction model generated in a system external to the question prediction system 10 to predict questions at the general meeting of shareholders.
  • the prediction unit 12 uses, for example, a question prediction model that learns the relationship between a predetermined feature amount extracted from the general meeting of shareholders related information and a question, to obtain a predetermined feature extracted from the general meeting of shareholders related information acquired by the acquisition unit 11. Predict questions based on quantity. When the general meeting of shareholders-related information consists only of data used for predicting questions, the prediction unit 12 does not need to perform the process of extracting feature amounts.
  • a question prediction model is generated for each shareholder meeting-related information. Generating for each shareholder meeting-related information means, for example, that different question prediction models are generated when predicting questions using financial information and when predicting questions using news. be.
  • the predetermined feature amount refers to some value indicating company information. When the general meeting of shareholders related information is character information, the predetermined feature amount is a value obtained by converting the character information indicating information about the company into a numerical value. The content and format of the predetermined feature amount are not particularly limited.
  • the prediction unit 12 uses the data of the financial information as input data and uses a question prediction model to predict questions to be asked at the general meeting of shareholders.
  • the prediction unit 12 extracts, for example, data set in a question prediction model from financial information as a predetermined feature amount.
  • the prediction unit 12 receives the extracted predetermined feature amount as an input and predicts a question using a question prediction model.
  • the prediction unit 12 uses, as predetermined feature amounts, for example, sales, operating income, ordinary income, net income, and operating income ratio in the latest accounting period. , ordinary income ratio, capital investment amount, R&D expenses, and year-on-year changes in each item are extracted as feature quantities. Items of financial information are not limited to the above examples.
  • the prediction unit 12 predicts the probability that the question will be asked along with the question. Then, the prediction unit 12 regards a question whose probability of being asked is equal to or higher than a predetermined criterion as a prediction result of the question to be asked at the shareholders' meeting. For example, the prediction unit 12 calculates the degree of similarity between a predetermined feature amount extracted from the general meeting of shareholders related information at the time of prediction and a predetermined feature amount extracted from the general meeting of shareholders related information associated with the question asked in the past. be the probability that the question is asked. The prediction unit 12 may use the questions from the top to the preset order with the highest probability of being asked as the prediction results of the questions to be asked at the general meeting of shareholders.
  • the prediction unit 12 extracts preset words or frequently used words from the words used in the news. Then, the prediction unit 12 uses the extracted words as input and predicts the question using the question prediction model.
  • the prediction unit 12 extracts words from news by a well-known natural language processing method.
  • News used for prediction is input to the terminal device 20 by the operator's operation, for example.
  • the news used for the prediction is, for example, an article about a company hosting a general meeting of shareholders that has been viewed more than a reference number of times.
  • News about the company hosting the general meeting of shareholders includes, for example, news with the name of the company hosting the general meeting of shareholders, news about the products or services handled by the company hosting the general meeting of shareholders, and the industry to which the company hosting the general meeting of shareholders belongs. This is news about. Examples of news about companies hosting shareholder meetings are not limited to the above.
  • the prediction unit 12 may extract a news field, use the extracted field as an input, and predict a question using a question prediction model.
  • the prediction unit 12 acquires, for example, statistical data of words contained in the news.
  • the prediction unit 12 extracts, for example, news fields corresponding to words whose number of appearances or frequency is set in advance and is equal to or greater than a reference value in the statistical data.
  • the acquisition unit 11 uses the extracted news field as an input, for example, and predicts a question using a question prediction model.
  • the relationship between words used in news and corresponding fields is set in advance. Areas of news are, for example, economics, politics, management, corporate trends, securities information and human resources.
  • the news fields may be product types, service types, and industries. The field of news is not limited to the above examples.
  • the prediction unit 12 may predict the type of question when predicting a question using a question prediction model with words contained in news or statistical data of words as input.
  • the types of questions are classifications when questions are divided into groups, such as questions about financial conditions, questions about investments, questions about priority areas, and questions about personnel measures. The types of questions are not limited to the above examples.
  • the prediction unit 12 may use the predicted question type as the question prediction result. When predicting the question type, the prediction unit 12 may generate a question sentence by replacing part of the predicted question with the word used as the input for the prediction.
  • the prediction unit 12 uses question sentence data in which the position of a word to be replaced in advance is used to replace the word, thereby generating a question sentence. Generate.
  • the prediction unit 12 predicts questions in the same way as news.
  • the prediction unit 12 may predict questions using different question prediction models for each prediction target.
  • the prediction unit 12 predicts questions using different question prediction models for, for example, prediction of questions about financial information and prediction of questions about news.
  • the prediction unit 12 may predict questions using different question prediction models for each company attribute or industry type.
  • the attributes of a company are, for example, the type of business, size, years of existence, or shareholder composition of the company that holds the shareholders' meeting.
  • Corporate attributes are not limited to the above.
  • the question prediction model may be selected based on any of the attributes described above, or may be selected based on a combination of two or more attributes.
  • the prediction unit 12 predicts a question using, for example, a question prediction model whose selection result is input to the terminal device 20 by the operator's operation.
  • the prediction unit 12 may generate a question prediction result by adding an answer example to the question.
  • the prediction unit 12 for example, extracts an answer example given to the same question from a database.
  • the prediction unit 12 also generates a prediction result by, for example, associating a question with an example answer extracted from a database.
  • a database of example answers given to the question is generated in advance and stored in the storage unit 15 .
  • the prediction unit 12 predicts questions that are likely to be asked at the general meeting of shareholders from information related to the general meeting of shareholders. There may be.
  • the prediction unit 12 may, for example, use a question prediction model to search for questions that are highly similar to sentences or words included in the general meeting of shareholders related information, and use the questions that have a high degree of similarity as prediction results of questions at the general meeting of shareholders. .
  • the prediction unit 12 may generate a prediction result by associating an answer associated with the question with the question.
  • the prediction model generation unit 13 generates a question prediction model that has learned the relationship between information related to the general meeting of shareholders and questions at the general meeting of shareholders.
  • the prediction model generation unit 13, for example, learns the relationship between financial information data and questions and generates a question prediction model.
  • the predictive model generating unit 13 learns the relationship between the financial information of the company that holds the general meeting of shareholders for several years and the question, and generates a question predictive model.
  • the predictive model generation unit 13 may learn the relationship between the financial information of a plurality of companies and the questions asked at the shareholders meeting of each company, and generate the question predictive model.
  • the shareholders' meeting-related information used by the prediction model generating unit 13 is not limited to this example.
  • the prediction model generation unit 13 learns the relationship between the predetermined feature amount extracted from the general meeting of shareholders related information and the question at the general meeting of shareholders, and receives the predetermined feature amount extracted from the general meeting of shareholders related information as an input. Generate a question prediction model that predicts questions in A feature amount to be extracted from the general meeting of shareholders-related information is set by, for example, an operator.
  • the prediction model generation unit 13 learns the relationship between a predetermined feature amount extracted from the general meeting of shareholders related information and the identifier assigned to the question, and predicts the probability of being asked for each question indicated by the identifier. Generate predictive models.
  • An identifier is information that identifies a question. For example, the same identifier is assigned to the same question among the questions used to generate the question prediction model. Identical questions may include similar questions.
  • the prediction model generation unit 13 learns the relationship between a predetermined feature amount extracted from at least one of financial information of multiple years and financial information of multiple companies, and the identifier assigned to the question, and the identifier indicates Generate a question prediction model that predicts the probability of being asked a question for each question.
  • Predetermined feature amounts extracted from financial information are set in advance.
  • the predetermined feature amount used to generate the question prediction model and the item of the predetermined feature amount used to predict the question using the question prediction model are the same for each question prediction model.
  • the prediction model generation unit 13 extracts items of financial information set in advance as feature amounts, learns the relationship between the extracted feature amounts and questions, and generates a question prediction model.
  • the prediction model generation unit 13 generates, for example, a question prediction model that predicts questions that are likely to be asked with respect to the extracted feature amount.
  • the prediction model generation unit 13 extracts preset words or words with high frequency from the words used in the news and extracts them. Generate a question prediction model that learns the relationship between words and questions. Word extraction from news is performed by well-known natural language processing methods.
  • the prediction model generation unit 13 may generate a question prediction model by extracting news fields and learning the relationship between the extracted fields and questions.
  • the prediction model generation unit 13 may acquire statistical data of news and generate a question prediction model by learning the relationship between the appearance frequency of each word and questions.
  • the prediction model generation unit 13 when prediction of questions is performed using different question prediction models for each company attribute or industry, the prediction model generation unit 13 generates each question prediction model using data for each prediction target. For example, when a different question prediction model is used for each type of business of a company, the prediction model generating unit 13 generates a question prediction model using the information related to the general meeting of shareholders collected for each type of business of the company and question data. do.
  • the prediction model generation unit 13 stores the generated question prediction model in the storage unit 15. When a plurality of question prediction models are generated, the prediction model generation unit 13 adds an identifier to each question prediction model, for example, and stores the question prediction models in the storage unit 15 .
  • the prediction model generation unit 13 may relearn the question prediction model using the questions asked at the shareholders' meeting.
  • the predictive model generating unit 13 updates the question predictive model by learning the relationship between, for example, the shareholders' meeting-related information updated or published after the last shareholders' meeting and the questions asked at the shareholders' meeting. When re-learning is performed, the prediction model generation unit 13 updates the question prediction model stored in the storage unit 15 .
  • the prediction model generation unit 13 generates a question prediction model by, for example, deep learning using a neural network.
  • the learning algorithm used to generate the question prediction model is not limited to the above example.
  • the output unit 14 outputs the result of prediction.
  • the output unit 14 outputs the question predicted by the prediction unit 12 to the terminal device 20 as a prediction result.
  • the output unit 14 outputs the prediction results in descending order of the possibility that the question will be asked.
  • the output unit 14 outputs questions in descending order of the probability that the question will be asked in the prediction results of the prediction unit 12 .
  • the output unit 14 may output the prediction result of the question for each field of the question.
  • the output unit 14 may output the prediction result to a display device (not shown) connected to the question prediction system 10 .
  • the output unit 14 may output the prediction result to a printing device connected via a network.
  • FIG. 3 is a diagram showing an example of the display screen of the question prediction result.
  • FIG. 3 shows prediction results of questions as prediction questions.
  • the expected questions include "Do you have any plans to make new investments in light of the favorable business performance?" Please tell us the specific measures for continuous growth.” are displayed.
  • the output unit 14 further outputs an answer example to the question.
  • FIG. 4 is a diagram showing an example of a display screen when an answer example is displayed on the display screen of the prediction result of the question.
  • an answer example to each question is shown together with the prediction result of the question similar to that of FIG.
  • the prediction result of the question similar to that of FIG.
  • the example of Figure 4 for example, in response to the forecast question, "Do you have any plans to make new investments in light of the favorable business performance?" .” is associated and displayed.
  • one answer example is shown for one question, but a plurality of answer examples for one question may be displayed.
  • the storage unit 15 stores data used for predicting questions.
  • the storage unit 15 stores the shareholders meeting-related information acquired by the acquisition unit 11 when the questions are predicted.
  • the storage unit 15 stores the result of question prediction by the prediction unit 12 .
  • the storage unit 15 stores the general meeting of shareholders-related information and question data used for generating the question prediction model acquired by the acquisition unit 11 when the question prediction model is generated.
  • the storage unit 15 stores the question prediction model generated by the prediction model generation unit 13 .
  • the terminal device 20 acquires the prediction result of the question from the question prediction system 10.
  • the terminal device 20 displays the prediction result of the question on a display device (not shown).
  • the terminal device 20 acquires the input result of the question prediction model selection result input by the operator's operation.
  • the terminal device 20 outputs the acquired selection result of the question prediction model to the question prediction system 10 .
  • a plurality of terminal devices 20 may be connected to the question prediction system 10.
  • a question prediction model to be used for prediction may be set in advance for each user logged in to the question prediction system 10 or for each terminal device 20 .
  • the terminal device 20 for example, a personal computer is used.
  • the terminal device 20 may be a smart phone or a tablet computer.
  • the terminal device 20 is not limited to the above example.
  • FIG. 5 is a diagram showing an example of an operation flow when predicting a question using a question prediction model in the question prediction system 10. As shown in FIG. 5
  • the acquisition unit 11 acquires information related to the company holding the general meeting of shareholders as information related to the general meeting of shareholders (step S11).
  • the acquiring unit 11 acquires information related to the general meeting of shareholders from the terminal device 20, for example.
  • the acquisition unit 11 stores the acquired information related to the general meeting of shareholders in the storage unit 15 .
  • the acquisition unit 11 After acquiring the information related to the general meeting of shareholders, the acquisition unit 11 acquires information on the selection result of the question prediction model used for question prediction (step S12). The acquisition unit 11 acquires, from the terminal device 20, information on the selection result of the question prediction model input to the terminal device 20 by the operator's operation, for example.
  • steps S11 and S12 is not particularly limited. Further, when the question prediction model used for prediction is set in advance, the acquisition unit 11 does not need to acquire the selection result of the question prediction model. For example, when a question prediction model used for prediction is set in advance for each user logged in to the question prediction system 10 or for each terminal device 20, the acquisition unit 11 acquires the selection result of the question prediction model. may not be performed.
  • the prediction unit 12 uses the selected question prediction model to predict questions based on the general meeting of shareholders related information (step S13).
  • the prediction unit 12 inputs the information related to the general meeting of shareholders acquired by the acquisition unit 11 as input data for the selected question prediction model, and predicts questions to be asked at the general meeting of shareholders.
  • the prediction unit 12 predicts the question using the question prediction model set in advance.
  • the prediction unit 12 extracts the feature amount set in the question prediction model from the information related to the general meeting of shareholders, and predicts the questions to be asked at the general meeting of shareholders from the extracted feature amount.
  • the output unit 14 When the question is predicted, the output unit 14 outputs the prediction result to the terminal device 20 (step S14).
  • the output unit 14 outputs, to the terminal device 20, the questions predicted by the prediction unit 12 that will be asked at the general meeting of shareholders as prediction results.
  • the output unit 14 associates the question with the answer example and outputs the prediction result to the terminal device 20 .
  • the terminal device 20 After obtaining the prediction result, the terminal device 20 displays the prediction result on the display device.
  • the output unit 14 associates the answer example with the question and displays it as the prediction result.
  • the terminal device 20 may output the prediction result to a device other than the display device.
  • the terminal device 20 may output the prediction result to, for example, a printing device.
  • FIG. 6 is a diagram showing an operational flow when the question prediction system 10 generates a question prediction model.
  • the acquisition unit 11 acquires information related to the general meeting of shareholders and data of questions at the general meeting of shareholders (step S21). For example, the acquisition unit 11 acquires from the terminal device 20 the information related to the general meeting of shareholders input to the terminal device 20 by the operator's operation and the data of the questions at the general meeting of shareholders. When a plurality of question prediction models are generated, the acquisition unit 11 acquires the general meeting of shareholders related information corresponding to the question prediction models to be generated and the data of the questions at the general meeting of shareholders.
  • the acquiring unit 11 After acquiring the shareholders' meeting-related information and question data, the acquiring unit 11 stores the acquired shareholders' meeting-related information and question data in the storage unit 15 .
  • the prediction model generating unit 13 learns the relationship between the information related to the general meeting of shareholders and the questions at the general meeting of shareholders, and generates a question prediction model (step S22 ).
  • the prediction model generation unit 13 After generating the question prediction model, the prediction model generation unit 13 stores the generated question prediction model in the storage unit 15 (step S23).
  • the prediction model generation unit 13 verifies the accuracy of the question prediction model, for example, using information related to the general meeting of shareholders and data that is not used to generate the question prediction model, among the question data.
  • the predictive model generation unit 13 receives information related to the general meeting of shareholders as input and outputs questions predicted by the question predictive model.
  • the prediction model generation unit 13 acquires a comparison result between the predicted question and the past question.
  • the comparison result is input to the terminal device 20 by the operator's operation, for example.
  • the prediction model generation unit 13 determines that generation of the question prediction model has been completed. If the ratio determined to match in the comparison result is less than the reference, the prediction model generation unit 13 further learns, for example, the relationship between the general meeting of shareholders related information and the question data, and continues to generate the question prediction model. do.
  • the question prediction system 10 of the information processing system of the present embodiment predicts questions at the general meeting of shareholders based on the information related to the general meeting of shareholders using a question prediction model that has learned the relationship between the questions and the information related to the general meeting of shareholders. to output Therefore, by using the question prediction system 10, it is possible to reduce the amount of work required for the operator to analyze information and investigate past questions, and to predict questions to be asked at the general meeting of shareholders. Therefore, by using the question prediction system 100, it is possible to easily predict questions at the shareholders' meeting.
  • the question prediction system 10 can predict questions with higher precision by inputting shareholder meeting-related information for a period that is likely to be reflected in questions and predicting questions.
  • the tendency of questions may change depending on the company holding the general meeting of shareholders or the attributes of the company. For example, the tendency of questions asked at shareholders' meetings can change depending on whether the company is a newly established company or a mature company. Also, the tendency of questions asked at general shareholders' meetings may change depending on whether there are many individual shareholders or institutional investors. Therefore, the question prediction system 10 can predict questions with higher accuracy by predicting questions using a question prediction model generated according to a company or an attribute of a company. As described above, by using the question prediction system 100 of the present embodiment, it is possible to predict questions according to the situation of the company that holds the shareholders' meeting.
  • FIG. 7 is a diagram showing an example of the configuration of the question prediction system 100 of this embodiment.
  • the question prediction system 100 of this embodiment includes an acquisition unit 101, a prediction unit 102, and an output unit 103.
  • the acquisition unit 101 acquires information related to a company that holds a general meeting of shareholders as information related to the general meeting of shareholders.
  • the prediction unit 102 predicts questions at the general meeting of shareholders based on the information related to the general meeting of shareholders acquired by the acquisition unit 101 using a question prediction model that predicts questions at the general meeting of shareholders from information related to the general meeting of shareholders.
  • the output unit 103 outputs the result of prediction.
  • the acquisition unit 11 of the first embodiment is an example of the acquisition unit 101.
  • the acquiring unit 101 is one aspect of an acquiring unit.
  • the prediction unit 12 of the first embodiment is an example of the prediction unit 102 .
  • the prediction unit 102 is one aspect of prediction means.
  • the output unit 14 of the first embodiment is an example of the output unit 103 .
  • the output unit 103 is one aspect of output means.
  • FIG. 8 is a diagram showing an example of the operational flow of the question prediction system 100. As shown in FIG. 8
  • the acquisition unit 101 acquires information related to the company that holds the general meeting of shareholders as information related to the general meeting of shareholders (step S101).
  • the prediction unit 102 uses a question prediction model that predicts questions at the general meeting of shareholders from the information related to the general meeting of shareholders.
  • a question is predicted (step S102).
  • the output unit 103 outputs the prediction result (step S103).
  • the question prediction system 100 of the present embodiment predicts questions at the general meeting of shareholders based on the information related to the general meeting of shareholders using a question prediction model that has learned the relationship between information related to the general meeting of shareholders and questions, and outputs the result of the prediction. .
  • the question prediction system 100 can predict questions according to the situation of the company holding the general meeting of shareholders.
  • FIG. 9 shows an example of the configuration of a computer 200 that executes computer programs for performing processes in the question prediction system 10 of the first embodiment and the question prediction system 100 of the second embodiment.
  • the computer 200 includes a CPU (Central Processing Unit) 201 , a memory 202 , a storage device 203 , an input/output I/F (Interface) 204 and a communication I/F 205 .
  • CPU Central Processing Unit
  • the CPU 201 reads a computer program for each process from the storage device 203 and executes it.
  • the CPU 201 may be configured by a combination of multiple CPUs.
  • the CPU 201 may be configured by a combination of a CPU and another type of processor.
  • the CPU 201 may be configured by a combination of a CPU and a GPU (Graphics Processing Unit).
  • the memory 202 is composed of a DRAM (Dynamic Random Access Memory) or the like, and temporarily stores computer programs executed by the CPU 201 and data being processed.
  • the storage device 203 stores computer programs executed by the CPU 201 .
  • the memory device 203 is configured by, for example, a nonvolatile semiconductor memory device. Other storage devices such as a hard disk drive may be used as the storage device 203 .
  • the input/output I/F 204 is an interface that receives input from the operator and outputs display data and the like.
  • a communication I/F 205 is an interface that transmits and receives data to and from a terminal device or another information processing device. Further, the terminal device 20 of the first embodiment can also have the same configuration.
  • Each process in the question prediction system 10 of the first embodiment and the question prediction system 100 of the second embodiment may be performed by multiple computers connected via a network. For example, processing related to question prediction and processing related to question prediction model generation may be performed on different computers.
  • the storage unit of the question prediction system 10 may be provided in a storage device connected via a network or a storage device managed by a server connected via a network.
  • the computer program used to execute each process can also be stored in a recording medium and distributed.
  • a recording medium for example, a magnetic tape for data recording or a magnetic disk such as a hard disk can be used.
  • an optical disc such as a CD-ROM (Compact Disc Read Only Memory) can be used.
  • a nonvolatile semiconductor memory device may be used as a recording medium.
  • a question prediction system comprising: output means for outputting prediction results;
  • the prediction means uses the question prediction model that has learned the relationship between a predetermined feature amount extracted from the general meeting of shareholders related information and the question, and uses the question prediction model to learn the predetermined Predicting the question based on the features of The question prediction system according to appendix 1.
  • the acquisition means acquires information related to the general meeting of shareholders during a period that is likely to be reflected in questions at the general meeting of shareholders.
  • the question prediction system according to appendix 1 or 2.
  • the acquisition means acquires the information related to the general meeting of shareholders updated or published after the previous general meeting of shareholders.
  • the prediction means predicts the question using the question prediction model generated for each company that holds a general meeting of shareholders or for each attribute of the company. 5.
  • the question prediction system according to any one of Appendices 1 to 4.
  • the shareholder meeting-related information is at least one of financial information, news, external announcements, or posts about companies,
  • the question prediction system according to any one of Appendices 1 to 6.
  • the output means further outputs an example answer to the predicted question, The question prediction system according to any one of Appendices 1 to 7.
  • the output means outputs the result of the prediction in descending order of the possibility that the question will be asked at the general meeting of shareholders. 9.
  • the question prediction system according to any one of Appendices 1 or 8.
  • Appendix 10 10. The question prediction system according to any one of appendices 1 to 9, further comprising prediction model generation means for generating a question prediction model that learns the relationship between the information related to the general meeting of shareholders and the questions at the general meeting of shareholders.
  • the predictive model generating means re-learns the relationship between the general meeting of shareholders related information and the questions asked at the general meeting of shareholders, and updates the question prediction model.
  • the question prediction system according to appendix 10.
  • [Appendix 12] Acquire information related to the company holding the general meeting of shareholders as information related to the general meeting of shareholders, Predicting questions at the general meeting of shareholders based on the obtained information related to the general meeting of shareholders using a question prediction model that predicts questions at the general meeting of shareholders from information related to the general meeting of shareholders, output the result of the prediction, Question prediction method.
  • Appendix 13 A process of acquiring information related to a company that holds a general meeting of shareholders as information related to the general meeting of shareholders; A process of predicting questions at the general meeting of shareholders based on the acquired information related to the general meeting of shareholders using a question prediction model that predicts questions at the general meeting of shareholders from information related to the general meeting of shareholders; A program recording medium for recording a process of outputting prediction results and a question prediction program for causing a computer to execute .

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • Marketing (AREA)
  • Game Theory and Decision Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Development Economics (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

Ce système de prédiction de question comprend une unité d'acquisition, une unité de prédiction et une unité de sortie. L'unité d'acquisition acquiert des informations relatives à une société qui va tenir une réunion d'actionnaires, en tant qu'informations relatives à une réunion d'actionnaires. L'unité de prédiction utilise un modèle de prédiction de questions pour prédire des questions aux réunions d'actionnaires, à partir d'informations relatives à une réunion d'actionnaires, de façon à prédire des questions lors d'une réunion d'actionnaires, sur la base des informations relatives à des réunions d'actionnaires, acquises par l'unité d'acquisition. L'unité de sortie délivre le résultat de prédiction.
PCT/JP2021/047466 2021-12-22 2021-12-22 Système de prédiction de question, procédé de prédiction de question et support d'enregistrement de programme WO2023119460A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/JP2021/047466 WO2023119460A1 (fr) 2021-12-22 2021-12-22 Système de prédiction de question, procédé de prédiction de question et support d'enregistrement de programme

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/JP2021/047466 WO2023119460A1 (fr) 2021-12-22 2021-12-22 Système de prédiction de question, procédé de prédiction de question et support d'enregistrement de programme

Publications (1)

Publication Number Publication Date
WO2023119460A1 true WO2023119460A1 (fr) 2023-06-29

Family

ID=86901683

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2021/047466 WO2023119460A1 (fr) 2021-12-22 2021-12-22 Système de prédiction de question, procédé de prédiction de question et support d'enregistrement de programme

Country Status (1)

Country Link
WO (1) WO2023119460A1 (fr)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2003085355A (ja) * 2001-09-12 2003-03-20 Nec Corp 会社支配支援システム、株主総会情報処理装置、株主権行使支援サーバ、会社支配支援方法及び制御プログラム
JP2020135434A (ja) * 2019-02-20 2020-08-31 国立大学法人一橋大学 企業情報処理装置、企業のイベント予測方法及び予測プログラム
JP2021089655A (ja) * 2019-12-05 2021-06-10 学校法人明治大学 学習モデル構築装置、学習モデル構築方法及びコンピュータプログラム

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2003085355A (ja) * 2001-09-12 2003-03-20 Nec Corp 会社支配支援システム、株主総会情報処理装置、株主権行使支援サーバ、会社支配支援方法及び制御プログラム
JP2020135434A (ja) * 2019-02-20 2020-08-31 国立大学法人一橋大学 企業情報処理装置、企業のイベント予測方法及び予測プログラム
JP2021089655A (ja) * 2019-12-05 2021-06-10 学校法人明治大学 学習モデル構築装置、学習モデル構築方法及びコンピュータプログラム

Similar Documents

Publication Publication Date Title
Zolas et al. Advanced technologies adoption and use by us firms: Evidence from the annual business survey
Hoque et al. Adoption of information and communication technology for development: A case study of small and medium enterprises in Bangladesh
CN110070391B (zh) 数据处理方法、装置、计算机可读介质及电子设备
US11954577B2 (en) Deep neural network based user segmentation
Van de Ven et al. Designing new business startups: Entrepreneurial, organizational, and ecological considerations
Liao et al. Venture gestation paths of nascent entrepreneurs: Exploring the temporal patterns
US20160196587A1 (en) Predictive modeling system applied to contextual commerce
Lan et al. Individual investment decision behaviors based on demographic characteristics: Case from China
CN104115178A (zh) 基于新闻和情绪分析来预测市场行为的方法和系统
Mau et al. Forecasting the next likely purchase events of insurance customers: A case study on the value of data-rich multichannel environments
Ahmed et al. Understanding the business value creation process for business intelligence tools in the UAE
Darwiesh et al. Business intelligence for risk management: A review
Ponzoa et al. EU27 and USA institutions in the digital ecosystem: Proposal for a digital presence measurement index
Bishnoi et al. Impact of AI and COVID-19 on manufacturing systems: An Asia Pacific Perspective on the two Competing exigencies
Rana et al. Emerging Technologies of Big Data in the Insurance Market
CN114240322A (zh) 业务处理方法、装置、存储介质和电子设备
Becker et al. National culture characteristics for managing corporate reputation and brand image using social media
Rouhani et al. Meta-synthesis of big data impacts on information systems development
WO2023119460A1 (fr) Système de prédiction de question, procédé de prédiction de question et support d'enregistrement de programme
Wilczek et al. Transforming the value chain of local journalism with artificial intelligence
Gerasimenko Digital strategy implementation in marketing: New performance and risks
Singh et al. Structural relation of export assistance programme and export performance determinants: A study of handloom industry
Al Hoderi Digital Economy: Definition, Advantages, Disadvantages
Yaghoobi et al. Identification and Ranking of Business Intelligence Components Using the Fuzzy TOPSIS Technique
Vassilev et al. What’s in a job? Measuring skills from online job adverts

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21968897

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE