Article creating system, article creating device, article creating method, and computer program

Abstract

To cause the contents of an article automatically created for a published account settlement, etc., to be focused on characteristic matters contained in the published contents.SOLUTION: An article creating device 20 automatically creates an article upon obtainment of, by an information DB system 10, company information like an account settlement distributed from a stock exchange system 2. The article creating device 20 compares numerical value information on each field contained in announced business performance contents with numerical value information on each field stored in a past account settlement DB 16 of the information DB system 10 and a business performance estimating DB 17 thereof, specifies a field that has a remarkable difference on the basis of the comparison result, and utilizes the specified field as the heading or the subject of the article, thereby automatically creating the article that has an emphasized characteristic matter contained in the announced contents.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an article generation system, article generation device, article generation method, and computer program for automatically generating articles, and in particular, to an article generation system, article generation device, article generation method, and computer program that compares the content of announcements regarding each company's performance with performance forecasts, etc., to identify items highly relevant to performance trends, and automatically generate articles that individually edit the performance status of each company based on the identified items. [Background technology]

[0002] Automatic generation of texts such as articles has been practiced for some time. Patent Document 1 listed below discloses generating a text report into which numbers or data values ​​are inserted using an appropriate template based on the type of event that occurs (see paragraphs 0023 to 0029 and Figure 1B of Patent Document 1). As an example, a report on a company's fourth-quarter profits is shown (see paragraphs 0031 and 0032 and Figure 4 of Patent Document 1).

[0003] Furthermore, Patent Document 2 below discloses that an explanatory text about a stock brand designated by a user is generated from stock price data, analysis of the stock price data, and news data (see paragraphs 0022 to 0053, and Figures 1 and 3, etc., of Patent Document 2). According to the disclosure of Patent Document 2, examples of explanatory texts generated by Patent Document 2 are like explanatory texts K1 and K2 shown in Figures 13(a) and 13(b).

[0004] Furthermore, artificial intelligence (AI) is now being used to automatically create articles. For example, Non-Patent Document 1 below discloses that when a listed company discloses (announces) financial statements on TDnet (a timely disclosure information viewing service) operated by the Tokyo Stock Exchange, an artificial intelligence (AI) summarizes the financial statements into three paragraphs: "Performance," "Factors," and "Outlook." The AI ​​then creates a text by applying the published figures to a template, and also describes the reasons for changes in sales and profits compared to the same period last year. Examples of articles created using Non-Patent Document 1 are shown in Figures 14(a) and 14(b). Figure 14(a) shows article K10 about listed company A, and Figure 14(b) shows article K11 about listed company B.

[0005] As another example of article creation using artificial intelligence (AI), the following non-patent document 2 discloses that in addition to the above-mentioned "TDnet," points contained in information from "EDINET," the Financial Services Agency's electronic disclosure system for disclosure documents, are provided as news (QUICK AI breaking news). [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Special Publication No. 2004-526264 [Patent Document 2] Japanese Patent Application Laid-Open No. 2001-351011 [Non-Patent Document 1] IT Pro - "Nikkei's AI reporter creates manuscripts in just 10 seconds, shocking the "financial results summary" 3D Motion Human Anatomy: The world's first 3D human anatomy app that reproduces the movements and morphology of a living human being," [online], [searched July 3, 2017], Internet<URL:http: / / itpro.nikkeibp.co.jp / atcl / column / 17 / 040400118 / 041200003 / > [Non-patent document 2] @Press - "Automatic Analysis News 'QUICK AI Breaking News' Release ~ Instantly Detecting Stock Price Impact ~" [online] [Retrieved July 3, 2017], Internet<URL:https: / / www.atpress.ne.jp / news / 116363> Summary of the Invention [Problem to be solved by the invention]

[0007] The report generated by Patent Document 1 is a template into which numbers or data values ​​are inserted, so the report content is uniform, making it difficult to grasp what important items, etc. that affect stock prices, etc., are, and the overall evaluation is difficult to understand at a glance. Similarly, the explanatory text generated by Patent Document 2 is limited to formulaic sentences describing the rise or fall of stock prices, etc., for the stock names specified by the user, making it difficult to grasp important items, etc. related to corporate performance, and also has the problem that the content is limited to an explanation of stock prices.

[0008] Furthermore, the articles created based on Non-Patent Document 1 are basically formulaic sentences in which the numerical values ​​contained in the information disclosed on "TDnet" are fitted into a template. As shown in articles K10 and K11 in Figures 14(a) and 14(b), the articles simply list multiple items and numerical values ​​related to the items, which makes it difficult for investors and others to understand at a glance what the important items are when making investment decisions, what the overall evaluation is, etc.

[0009] Furthermore, the news created by Non-Patent Document 2 basically only uses information disclosed on "TDnet" and does not compare it with corporate earnings forecasts, etc., resulting in the problem of uniform content.

[0010] The present invention has been made in consideration of the above circumstances, and aims to provide an article generation system, article generation device, article generation method, and computer program that compares the numerical information contained in announcements of financial results, dividends, revisions, etc. of each company with performance forecasts, etc., to identify items that have an impact on a company's performance, stock price, etc., and applies such items to the headline or subject of the article, thereby automatically generating articles that make it easy to convey the matters that investors and others are paying attention to.

[0011] Another object of the present invention is to provide an article generation system, article generation device, article generation method, and computer program that identify items related to business performance, etc. included in the announced content for each announcement, edit the article centered on those identified items, and automatically generate articles with individual content tailored to the business performance, etc. of each company.

[0012] Furthermore, the present invention aims to provide an article generation system, an article generation device, an article generation method, and a computer program that can help investors and the like make quick decisions about a company's announced performance, etc. by presenting an overall evaluation of the company's announced performance, etc. [Means for solving the problem]

[0013] In order to solve the above problem, the present invention provides an article generation system that includes an article generation device that automatically generates company information articles based on company information distributed from an information distribution source, the company information distributed from the information distribution source including numerical information corresponding to a plurality of items related to the company's performance, and an information database system that receives the company information distributed from the information distribution source and stores predicted company information including predicted numerical information corresponding to the plurality of items, the article generation device including: means for detecting that new company information distributed from the information distribution source has been received by the information database system; means for, when the information database system detects that new company information has been received by the information database system, acquiring the new company information and predicted company information for the same company as the company related to the new company information from the information database system; means for comparing the numerical information corresponding to the plurality of items included in the acquired new company information with predicted numerical information corresponding to the plurality of items included in the acquired predicted company information, item by item; means for identifying, from the compared numerical information, the item related to numerical information with the largest value corresponding to the difference between the numerical information; and means for describing the information corresponding to the identified item in the headline or subject of the company information article.

[0014] The present invention is also characterized in that the information database system stores company information received from an information distribution source as past company information, and the article generation device, when detecting that new company information has been received by the information database system, is equipped with a means for acquiring, from the information database system, in addition to the new company information and the predicted company information, past company information of the same company as the company related to the new company information, and a means for comparing, for each item, numerical information corresponding to a plurality of items included in the new company information with predicted numerical information corresponding to a plurality of items included in the predicted company information and past numerical information corresponding to a plurality of items included in the acquired past company information.

[0015] Furthermore, the present invention is characterized in that the information database system stores average numerical information obtained by averaging numerical information corresponding to multiple items contained in the corporate information of multiple companies, and the article generation device comprises: a means for acquiring average numerical information in conjunction with the acquisition of new corporate information; a means for comparing the numerical information corresponding to multiple items contained in the acquired new corporate information with the acquired average numerical information for each item; a means for identifying, based on the comparison, the order of items with the largest value corresponding to the difference between the numerical information and the average numerical information; and a means for automatically generating a corporate information article by entering information related to each item in accordance with the identified order of the items.

[0016] Furthermore, the present invention is characterized in that the information database system comprises a means for storing theoretically calculated stock prices for each company's stock price, and the article generation device comprises a means for acquiring, upon acquisition of new company information, the theoretical stock prices of companies corresponding to the new company information, a means for acquiring the actual stock prices of each company to be distributed, and a means for automatically generating company information articles including the acquired theoretical stock prices and the actual stock prices of the same companies as the companies related to the theoretical stock prices.

[0017] The present invention is characterized in that the information database system comprises a means for storing investor forecast stock prices calculated by aggregating stock prices predicted by multiple investors for each company's stock price, and the article generation device comprises a means for acquiring, upon acquisition of new company information, investor forecast stock prices for companies corresponding to the new company information, a means for acquiring actual stock prices for each company that are distributed, and a means for automatically generating company information articles that include the acquired investor forecast stock prices and actual stock prices of the same companies as the companies related to the investor forecast stock prices.

[0018] The present invention also provides an article generation device that automatically generates company information articles based on company information distributed from an information distribution source, the article generation device receiving the company information distributed from the information distribution source and including numerical information corresponding to multiple items related to the company's performance, comprising: means for receiving the company information distributed from the information distribution source and detecting that new company information has been received by an external information database system that stores predicted company information including predicted numerical information corresponding to the multiple items; means for, when receiving new company information from the external information database system, acquiring the new company information and predicted company information for the same company as the company related to the new company information from the external information database system; means for comparing, item by item, the numerical information corresponding to the multiple items included in the acquired new company information with the predicted numerical information corresponding to the multiple items included in the acquired predicted company information; means for identifying, among the compared numerical information, the item related to numerical information with the largest value corresponding to the difference between the numerical information; and means for describing the information corresponding to the identified item in the headline or subject of the company information article.

[0019] Furthermore, the present invention is characterized in that an external information database system stores corporate information received from an information distributor as past corporate information, and when it detects that new corporate information has been received by the external information database system, it is equipped with a means for acquiring, from the external information database system, in addition to the new corporate information and the predicted corporate information, past corporate information of the same company as the company related to the new corporate information, and a comparison means for comparing, item by item, numerical information corresponding to multiple items included in the new corporate information with predicted numerical information corresponding to multiple items included in the predicted corporate information and past numerical information corresponding to multiple items included in the acquired past corporate information.

[0020] Furthermore, the present invention is characterized in that the comparison means compares the numerical information contained in the new company information with numerical information that is an extreme value in the past among the numerical information contained in the past company information.

[0021] The present invention is characterized in that when the plurality of items include important items and the numerical information included in the new company information is both larger and smaller than the numerical information of the comparison target, if the numerical information included in the new company information is larger than the numerical information of the comparison target in comparing the numerical information related to the important items, the invention is provided with an evaluation means for giving a positive overall evaluation to the new company information, and a means for attaching the positive overall evaluation to the company information article.

[0022] Furthermore, the present invention is characterized in that when the numerical information contained in the new company information is both larger and smaller than the numerical information of the comparison target, if the numerical information contained in the new company information is smaller than the numerical information of the comparison target in the comparison related to the important item, the evaluation means is configured to give the new company information a negative overall evaluation, and is provided with a means for attaching the negative overall evaluation to the company information article. Furthermore, the present invention is characterized in that the important item is ordinary profit.

[0023] Furthermore, the present invention is characterized in that the external information database system stores average numerical information obtained by averaging numerical information corresponding to multiple items contained in the corporate information of multiple companies, and the invention is equipped with a means for acquiring the average numerical information from the external information database system in conjunction with the acquisition of new corporate information, a means for comparing the numerical information corresponding to the multiple items contained in the acquired new corporate information with the acquired average numerical information for each item, and a means for designating the item associated with numerical information having the largest value corresponding to the difference between the numerical information and the average numerical information based on the comparison as the important item.

[0024] The present invention is characterized in that an external information database system stores average numerical information obtained by averaging numerical information corresponding to multiple items contained in the corporate information of multiple companies, and the invention includes a means for acquiring the average numerical information from the external information database system in conjunction with the acquisition of new corporate information, a means for comparing the numerical information corresponding to the multiple items contained in the acquired new corporate information with the acquired average numerical information for each item, a means for identifying the order of the items related to the numerical information to be compared based on the comparison in order of the largest value corresponding to the difference between the numerical information and the average numerical information, and a means for generating a corporate information article by entering information related to each item in accordance with the identified order of the items.

[0025] Further, the present invention provides an article generation method in which an article generation device automatically generates a company information article based on company information distributed from an information distribution source, the company information including numerical information corresponding to a plurality of items related to the company's performance, the article generation device comprising the steps of: receiving the company information distributed from the information distribution source; and detecting that new company information has been received by an external information database system that stores predicted company information including predicted numerical information corresponding to the plurality of items; when the article generation device detects that new company information has been received by the external information database system, acquiring the new company information and predicted company information of the same company as the company related to the new company information from the external information database system; comparing the numerical information corresponding to the plurality of items included in the acquired new company information with the predicted numerical information corresponding to the plurality of items included in the acquired predicted company information, item by item; identifying the item related to numerical information with the largest value corresponding to the difference between the numerical information among the compared numerical information; and describing the information corresponding to the identified item in the headline or subject of the company information article.

[0026] Furthermore, the present invention provides a computer program for causing a computer to automatically generate a company information article based on company information distributed from an information distribution source, the computer receiving the company information and storing predicted company information including predicted numerical information corresponding to the multiple items, the computer being configured to execute the following steps: receiving the company information distributed from the information distribution source and detecting that new company information has been received from an external information database system that stores predicted company information including predicted numerical information corresponding to the multiple items; when receiving new company information from the external information database system, acquiring the new company information and predicted company information for the same company as the company related to the new company information from the external information database system; comparing the numerical information corresponding to the multiple items included in the acquired new company information with the predicted numerical information corresponding to the multiple items included in the acquired predicted company information, item by item; identifying the item related to numerical information with the largest value corresponding to the difference between the numerical information among the compared numerical information; and including the information corresponding to the identified item in the headline or subject of the company information article.

[0027] In this invention, numerical information corresponding to multiple items contained in the new company information is compared with the forecast numerical information corresponding to multiple items contained in the forecast company information (performance forecast information) for each item, and the item associated with the numerical information with the largest value corresponding to the difference is identified, so that the item with the largest difference compared to the forecast is identified among the multiple items contained in the distributed company information.Then, by including that item in the headline or subject of the article, it becomes possible to provide an article that highlights the item that stands out in comparison with the forecast.

[0028] In the present invention, the numerical information corresponding to multiple items contained in new company information is compared with the prediction as well as with each piece of numerical information corresponding to multiple items contained in past company information, and the item associated with the numerical information corresponding to the value with the largest difference is identified.Therefore, of the multiple items contained in the company information to be distributed, the item with the largest difference compared to the prediction or past is identified, and this item is stated in the headline or subject of the article, making it possible to provide an article in which items that are distinctive in comparison with the prediction or past are highlighted.

[0029] In this invention, numerical information corresponding to multiple items contained in new company information is compared for each item with average numerical information obtained by averaging numerical information corresponding to multiple items contained in the company information of multiple companies, the order of items with the largest difference between the two is identified, and an article is generated by listing information related to each item in that order, resulting in article content that lists items with the largest differences from other companies in order, making it possible to provide articles that are structured to make it easy to understand the differences in performance, etc. of other companies that investors and others are paying close attention to.

[0030] In the present invention, an article is generated that includes a theoretically calculated theoretical stock price of a company and the actual stock price of that company, making it possible to provide an article that makes it easy for investors to understand the difference between the actual stock price and the theoretical stock price, which can be used as a reference for future investments.

[0031] In the present invention, an article is generated that includes an investor forecast stock price for a company, calculated by aggregating the stock prices predicted by multiple investors, and the actual stock price of the company. This makes it possible to provide articles that make it easy for investors to understand the difference between the forecast stock price, which represents the collective wisdom of many investors, and the actual stock price, and investors can use such articles as material for considering future investments.

[0032] In this invention, numerical information contained in new company information is compared with numerical information representing past extreme values ​​(maximum or minimum values), so it is possible to express in an article that a company's performance has reached a new high or low, and the current state of the company's performance can be clearly expressed in the context of a series of company performance trends. Note that the range of the past is preferably limited to a required period such as 10 years, 5 years, 3 years, 1 year, 6 months, 4 months, or 3 months, so as not to deviate too much from the current state.

[0033] In comparing each item, if the numerical information contained in the new corporate information is a mixture of values ​​that are both larger (positive) and smaller (negative) than the comparison target, it is generally difficult to evaluate whether the content of the newly released corporate information is good or not. However, in the present invention, important items are defined among multiple items, and when the comparison results are a mixture of positive and negative, as described above, making evaluation difficult, comparison is made by focusing on the important items. If the comparison result of the important items is positive, the overall evaluation of the newly released corporate information is made positive. Therefore, even when the content of the corporate information is a mix of good and bad items and difficult to evaluate, evaluation based on important items makes it possible to provide investors and others with a certain guideline that is easy to understand.

[0034] Furthermore, in the present invention, when the comparison results are a mixture of positive and negative as described above, if the comparison results in a negative comparison in the important items, the article will indicate that the overall evaluation of the new company information is negative. Therefore, even when the content of the company information is difficult to evaluate, it is possible to provide investors and the like with content that allows them to see the overall evaluation at a glance by evaluating it from the perspective of the important items.

[0035] Furthermore, in the present invention, current profit is set as the key item, and since current profit is generally the most significant indicator of a company's performance, by setting current profit as the key item, it becomes easier to appropriately evaluate the overall newly announced company information based on current profit.

[0036] In the present invention, numerical information corresponding to a plurality of items contained in newly released corporate information is compared for each item with average numerical information obtained by averaging the numerical information corresponding to a plurality of items contained in the corporate information of a plurality of companies, and the item with the largest difference between the two is designated as the important item, thereby enabling an overall evaluation of the new corporate information to be made with emphasis on the item that is most different from other companies. [Effects of the Invention]

[0037] In the present invention, among multiple items contained in new company information to be distributed, the item that differs the most from the forecast is included in the headline or subject of the article, so that investors and the like can be provided with automatically generated articles that highlight distinctive items in comparison with the forecast.

[0038] In addition to the forecast, the system also compares each item with the numerical information for multiple items contained in past company information, and identifies the item associated with the numerical information corresponding to the value with the largest difference.By comparing over a wide time range from the past to the future, the item with the largest difference is identified, and this item is included in the headline or subject of the article, making it possible to provide an article that highlights distinctive items in the company's performance when viewed both in the past and future.

[0039] Furthermore, in the present invention, if any of the items included in the new company information to be distributed reaches a new record high or record low, this fact is stated in the headline or subject of the article, thereby clearly conveying to investors and the like that the company's current performance is at its peak. Furthermore, in the present invention, the article contents are arranged in order of the items that are most different from other companies, so that an article can be provided that is structured to make it easier for investors and the like to grasp the prominent contents when compared with other companies that they pay close attention to.

[0040] In the present invention, an article is generated that includes a theoretically calculated theoretical stock price of a company and the actual stock price of that company, so that an article that allows comparison of the actual stock price with the theoretical stock price can be provided as investment decision material. Furthermore, in the present invention, an article is generated that includes investor-forecasted stock prices calculated by aggregating stock prices predicted by multiple investors, as well as actual stock prices, making it possible to provide an article that allows comparison of forecasted stock prices based on the collective intelligence of many investors with actual stock prices.

[0041] In the present invention, important items are defined among a plurality of items, and if the comparison results for each item are a mixture of positive and negative results, the article will indicate whether the overall evaluation of the newly announced corporate information is positive or negative based on the comparison results for the important items, so that the evaluation based on the important items can be presented as a certain guideline to investors, etc. Furthermore, in the present invention, by setting the important item to current profit, which clearly reflects the business performance of a company, it is possible to appropriately evaluate newly announced company information. Furthermore, the present invention allows for an overall evaluation of newly released company information, focusing on the items that are most different from other companies. [Brief explanation of the drawings]

[0042] [Figure 1] 1 is a schematic diagram illustrating an overall configuration of an article providing system including an article generation system according to an embodiment of the present invention. [Figure 2] FIG. 1 is a schematic diagram showing the internal configuration of an information DB system. [Figure 3] 2 is a block diagram showing the main internal configuration of the article generation device; FIG. [Figure 4] 10 is a diagram showing an example of the contents of a quarterly settlement table portion included in an evaluation criteria table. [Figure 5] 10 is a diagram showing an example of the contents of a full-year settlement table section included in an evaluation criteria table. [Figure 6] 1 is a flowchart showing a series of processing steps related to an article generating method. [Figure 7] FIG. 1 is a schematic diagram showing an example of a company information article automatically generated for the announced financial results of Company A. [Figure 8] FIG. 1 is a schematic diagram showing an example of a company information article automatically generated in response to the announced financial results of Company B. [Figure 9] (a) is a schematic diagram showing an example of a company information article automatically generated in response to the announcement of a revised earnings forecast for Company C, and (b) is a schematic diagram showing an example of a company information article automatically generated in response to the announcement of a revised earnings forecast for Company D. [Figure 10] FIG. 1 is a schematic diagram showing an example of an automatically generated company information article for Company E's announced dividend adjustment. [Figure 11] (a) is a schematic diagram showing an example of stock price information including Company A's theoretical stock price, (b) is a schematic diagram showing an example of stock price information including Company A's investor forecast stock price, and (c) is a schematic diagram showing an example of a comprehensive evaluation (overall evaluation) of Company A's announced financial results. [Figure 12] FIG. 1 is a schematic diagram showing an example of a visualized (illustrated) company information article. [Figure 13] 1(a) and 1(b) are schematic diagrams showing examples of explanatory text generated by the prior art. [Figure 14] (a) is a schematic diagram showing an example of an article automatically generated using conventional technology in response to the announced financial results of Company A, and (b) is a schematic diagram showing an example of an article automatically generated using conventional technology in response to the announced financial results of Company B. DETAILED DESCRIPTION OF THE INVENTION [Example]

[0043] 1 is a schematic diagram showing the overall configuration of an article providing system 1 including an article generation system 5 according to the present invention. The article providing system 1 automatically generates more sophisticated written articles than conventional ones using the article generation system 5, and makes them available to each user's user terminal T1, T2, T3, T4, etc. via a network. The article providing system 1 has a stock exchange system 2 that distributes various corporate information such as financial statements announced by each listed company as timely disclosure information in a predetermined standardized file format (for example, XBRL: eXtensible Business Reporting Language), and a stock price distribution system 3 that distributes fluctuating stock prices of each listed company as needed.

[0044] The article generation system 5 is composed of an information DB system 10 and an article generation device 20. The information DB system 10 is an information database system equipped with multiple types of DBs (databases) that store independently constructed data archives. In this embodiment, the information DB system 10 includes a current financial statement DB 15, a past financial statement DB 16, a performance forecast DB 17, a theoretical stock price DB 18, and an individual investor stock price forecast DB 19. The information DB system 10 periodically acquires various information (corresponding to the company information announced by listed companies) distributed by the stock exchange system 2. The article generation device 20 monitors (detects) the acquisition of such company information. The article generation device 20 automatically generates an article based on the announced company information when the information DB system 10 acquires the company information. The present invention will be described in detail below, focusing on the information DB system 10 and the article generation device 20 that constitute the article generation system 5.

[0045] 2 is a schematic diagram showing the internal configuration of the information DB system 10, in which a DB management device 11 is connected to each of the above-mentioned DBs 15 to 19 via internal connection lines 10a. The DB management device 11 manages and controls each of the DBs 15 to 19, and receives (acquires) information and data distributed (provided) from external systems, etc., sorts the received information, etc. according to its content, and stores it in each of the required DBs 15 to 19, and also reads out information, etc. stored in each of the DBs 15 to 19 in response to an external request, and transmits the read information, etc. to the requestor.

[0046] The DB management device 11 has an internal control unit 11a and a communication unit 11b, and the control unit 11a is essentially responsible for each of the processes of the DB management device 11 described above and controls each of these processes. The control unit 11a has an internal memory (storage unit) so that it can temporarily store information, data, etc. associated with the processes. The communication unit 11b communicates with an external network and each of the DBs 15 to 19 under the control of the control unit 11a, and communication by this communication unit 11b makes it possible to send and receive each of the information, data, etc. to and from each of the DBs 15 to 19.

[0047] The relevant financial statement DB 15 of the information DB system 10 is a database that stores company information released by each listed company distributed by the above-mentioned stock exchange system 2, in association with each company, and when the DB management device 11 receives and acquires company information distributed from the stock exchange system 2 (corresponding to the information distributor), it stores (stores) the acquired company information in the relevant financial statement DB 15. The company information distributed from the stock exchange system 2 is in the above-mentioned predetermined standardized file format (XBRL format), and the numerical information included in the XBRL format file is written in units of 1 million yen (for example, 10 billion yen is written as "10000"). Therefore, since it would be difficult to analyze the corporate information or generate documents at a later stage if the information were left as is, the DB management device 11 performs a conversion process of the numerical units for the numerical information contained in the acquired XBRL format corporate information when storing it in the relevant financial statements DB 15, and saves (stores) the converted content in the relevant financial statements DB 15 (for example, the numerical information of "10000" mentioned above is converted to the numerical unit "10000000000").

[0048] Furthermore, the DB management device 11 is configured to be able to externally detect the receipt of corporate information distributed from the stock exchange system 2, and when corporate information is received, it issues an information signal indicating which company's corporate information was received and when, and performs processing to transmit this issued information signal to the outside. The corporate information distributed by the stock exchange system 2 includes financial results (full-year, first half, second half, quarterly cumulative, quarterly, etc.) announced by each listed company, performance forecasts, performance forecast revisions, dividend forecasts, dividend forecast revisions, etc. This corporate information such as financial results, performance forecasts, performance forecast revisions, dividend forecasts, dividend forecast revisions, etc. includes numerical information (information indicating percentages such as %, numerical values ​​corresponding to monetary units, or unitless numerical values) corresponding to multiple items related to corporate performance, and is distributed to the information DB system 10 in accordance with the timely disclosure information viewing service operated by the stock exchange system 2.

[0049] If the corporate information is financial statements, the multiple items may include ordinary profit, operating profit, sales, net profit, etc., and the corporate information will include numerical information for each of these items for each company (numerical information indicating the numerical value of ordinary profit, numerical information indicating the numerical value of sales, numerical information indicating the numerical value of operating profit, numerical information indicating the numerical value of net profit, etc.) It should be noted that corporate information is generally distributed (announced) from the stock exchange system 2 after the market closes (at 3:00 PM) or around 3:30 PM, and the article generation system 5 of the present invention is capable of providing generated articles at approximately the same time as these announcement times.

[0050] In the case of revisions to corporate information (revisions to earnings forecasts, revisions to dividend forecasts, etc.), items corresponding to the subject of the revision will be included, including revisions to profit-related items such as revisions to operating profits, revisions to net profits, and figures (numerical information) indicating revisions to dividends.

[0051] The past financial statements DB 16 of the information DB system 10 is a database that stores past corporate information (past corporate information). The past corporate information is the archived and accumulated corporate information stored in the relevant financial statements DB 15. The storage process for the past financial statements DB 16 is linked to the storage process for the relevant financial statements DB 15. When the DB management device 11 stores corporate information in the relevant financial statements DB 15, the stored corporate information is also stored in the past financial statements DB 16. Therefore, only the most recently announced new corporate information is stored in the relevant financial statements DB 15, and outdated corporate information is deleted as needed (e.g., information released approximately two or three weeks after the announcement). On the other hand, once corporate information is stored in the past financial statements DB 16, it is stored as past corporate information in association with each company, making it a suitable database for checking past performance, etc. Note that the corporate information stored in the past financial statements DB 16 also contains numerical information in the XBRL format that has been converted.

[0052] Furthermore, the DB management device 11 calculates an average value for each of a plurality of items for the numerical information included in the past company information for each company stored in the past settlement DB 16, and stores the calculated average value for each item as average numerical information in the past settlement DB 16. This average value calculation process is performed by calculating the average value for all companies excluding one company. For example, assuming that companies A to Z are the companies from which company information is obtained, the calculation of the average value excluding company A is performed using the company information for companies B to Z, the calculation of the average value excluding company B is performed using the company information for companies A and C to Z, the calculation of the average value excluding company C is performed using the company information for companies A, B, and D to Z, and the calculation of the average value excluding company Z is performed using the company information for companies A to Y.

[0053] The DB management device 11 then stores the average value excluding Company A (the calculated average value) in the past financial statement DB 16 in association with Company A. Similarly, the average value excluding Company B (the calculated average value) is stored in association with Company B in the past financial statement DB 16, the average value excluding Company C (the calculated average value) is stored in association with Company C in the past financial statement DB 16, and the average value excluding Company Z (the calculated average value) is stored in association with Company Z in the past financial statement DB 16. By calculating average values ​​and storing them in association with each company in this manner, the stored average values ​​can be used when comparing the performance of a certain company with the average values ​​of other companies. Note that the DB management device 11 calculates such average values ​​at predetermined intervals (for example, monthly, such as every one month or three months, or annually, such as every six months, one year, three years, or five years) in accordance with the announcement of performance by each company. A plurality of average values ​​may be calculated and stored, such as a three-month average value, a six-month average value, a one-year average value, a three-year average value, or a five-year average value.

[0054] The performance forecast DB 17 of the information DB system 10 is a database that stores performance forecast information (equivalent to forecasted company information) including numerical information predicting the performance of each listed company, in association with each company. The performance forecast information to be stored is generated by the operating entity of the information DB system 10 or a forecasting company, etc., and, similar to the company information distributed by the stock exchange system 2, includes numerical information for forecasts based on multiple items related to the company's performance. When performance forecast information generated by a forecasting company's device or the like is sent to the information DB system 10, the DB management device 11 receives and acquires the performance forecast information and stores it in the performance forecast DB 17, distinguishing between companies. Note that this performance forecast information also conforms to the predetermined standardized file format described above, and the numerical units of the numerical information are common units (units whose first digit begins with the ones digit).

[0055] Furthermore, the theoretical stock price DB18 of the information DB system 10 is a database that stores theoretical stock prices (theoretical stock price information) indicating stock price values ​​calculated for each listed company using a formula based on a predetermined theory, etc., in association with each company. Such theoretical stock price information is calculated and generated as needed by, for example, the business entity that manages and operates the article generation device 20, the business entity that manages and operates the information DB system 10, or an external business entity, and when the theoretical stock price information is sent to the information DB system 10 from a device by such a business entity (a device that has the calculated theoretical stock price information), the DB management device 11 receives and acquires the theoretical stock price information and performs a process of storing it in the theoretical stock price DB18, distinguishing between each company.

[0056] Finally, the individual investor stock price forecast DB 19 of the information DB system 10 is a database that stores individual investor stock price forecast information, which indicates the stock prices forecast by multiple individual investors for each company. The individual investor stock price forecast information to be stored is provided by a business entity that compiles the individual investor's stock price forecasts. A device owned by this business entity compiles the individual investor's stock price forecasts and performs statistical processing (e.g., averaging) to generate stock price information based on the individual investor's forecasts (individual investor stock price forecast information including numerical information predicting the stock price of each company), and provides (transmits) the generated individual investor stock price forecast information to the information DB system 10, etc. Examples of business entities that compile such forecast stock prices include the business entity that manages and operates the article generation device 20, the business entity that manages and operates the information DB system 10, and external business entities. When the information DB system 10 receives individual investor stock price forecast information sent from the device of the business entity that compiles the forecast stock prices, the DB management device 11 stores the received individual investor stock price forecast information in the individual investor stock price forecast DB 19, correlating it with each company.

[0057] The DB management device 11 of the information DB system 10 manages and controls the DBs 15 to 19 described above, and when it receives various types of information about each company from outside the information DB system 10, it detects the content of the information and performs a process of storing (storing) the received information in one of the DBs 15 to 19 according to the content so that the information can be identified for each company. Furthermore, when the DB management device 11 of the information DB system 10 receives a request for information about a specific company (company information request) from an article generation device 20 that automatically generates articles, it reads out from each DB 15 to 19 various types of information (various types of information about the requested company) that are stored in association with the company, and performs a process of transmitting (providing) the information to the article generation device 20 that made the request (the various types of information are transmitted as files).

[0058] Specifically, when the DB management device 11 receives an information request for Company A, a listed company, the DB management device 11 reads the latest company information for Company A (company information file corresponding to the latest financial results, earnings forecast, earnings forecast revisions, dividend forecast, dividend forecast revisions, etc. distributed by the stock exchange system 2) from the relevant financial statements DB 15, reads past company information for Company A (past company information file) and information showing the average value of other companies excluding Company A (average numerical information) from the past financial statements DB 16, reads Company A's earnings forecast information (earnings forecast information file) from the earnings forecast DB 17, reads Company A's theoretical stock price information (theoretical stock price information file) from the theoretical stock price DB 18, and reads Company A's individual investor forecast stock price information (individual investor forecast stock price information file) from the individual investor stock price forecast DB 19, and transmits the various information for Company A that it has read (files of each information) to the article generation device 20. In this embodiment, the past corporate information read from the past financial statements DB 16 is information for the past five years (information on financial statements, business performance, dividends, etc. announced within the past five years), but depending on the system specifications, etc., appropriate past information within a range of about the past ten years may be applied. Also, the information DB system 10 corresponds to an external information database system for the article generation device 20.

[0059] 3 is a block diagram showing the main internal configuration of the article generation device 20, which is the central device of the present invention. The article generation device 20 of this embodiment is constructed using a general server computer (server device), but it is of course possible to construct a system by combining multiple server devices by performing distributed processing, etc. The article generation device 20 uses a unique program (AI: artificial intelligence) to automatically generate articles based on various information and data archived in the above-mentioned information DB system 10, triggered by the announcement of corporate information.

[0060] The article generation device 20 is configured by connecting various devices, etc., via internal connection lines 20h to an MPU 20a that performs overall control and various processing, and the various devices, etc., include a communication module 20b, a RAM 20c, a ROM 20d, an input interface 20e, an output interface 20f, and a memory unit 20g.

[0061] The communication module 20b is a communication device equivalent to a network connection module, and conforms to a required communication standard (e.g., a LAN module). The communication module 20b is connected to a network (including internal, external, or other communication networks) via required communication equipment (not shown, such as a router), enabling communication with the stock price distribution system 3, the information DB system 10 (which corresponds to an external information database system for the article generation device 20), and user terminals T1, T2, T3, etc. The stock price distribution system 3 distributes stock prices of each company during the hours the stock market is open, and the distributed stock prices are received by the communication module 20b, and the received distributed stock prices are sequentially stored in the RAM 20c or the memory unit 20g, etc.

[0062] The RAM 20c temporarily stores the contents, files, etc. associated with the processing of the MPU 20a, and the ROM 20d stores the programs, etc. that define the basic processing contents of the MPU 20a. The input interface 20e is connected to a keyboard 21, mouse, etc. that accepts operation instructions, etc. from the administrator, etc. of the article generating device 20. The output interface 20f is connected to a display 22 (display output device), and outputs the contents associated with the processing of the MPU 20a to the display 22 so that the administrator, etc. of the article generating device 20 can check the current processing contents, etc.

[0063] The storage unit 20g stores programs, tables, etc., and stores the OS program P1 and the article generation program P2 as programs, and the created article table 24 and the evaluation criteria table 25 as tables. Note that, in order to install the article generation program P2 into the storage unit 20g, it is conceivable to store the article generation program P2 in a storage medium such as an optical disk and install it into the storage unit 20g via the storage medium.

[0064] Among the programs stored in the memory unit 20g, the OS program P1 specifies various processes according to the operating system for the server computer, and the MPU 20a performs processes based on these specifications, allowing the article generation device 20 to perform each function as a server computer (server device).

[0065] The article generation program P2 stored in the memory unit 20g defines various processes for the MPU 20a to automatically generate articles as an artificial intelligence. In accordance with the provisions of the article generation program P2, the MPU 20a functions as a means for monitoring the reception of corporate information by the information DB system 10, analyzing and evaluating the corporate information acquired from the information DB system 10, and identifying article headlines, important items, and the order of explanation of the article content based on the results of the analysis and evaluation. To create sentences related to articles, the article generation program P2 also includes processing content according to a sentence creation engine. This sentence creation engine allows the MPU 20a to function as an AI (artificial intelligence) for sentence creation. Details of the article generation program P2 will be described below with reference to the flowchart shown in FIG. 6. Before that, the created article table 24 and the evaluation criteria table 25 will be described.

[0066] The created article table 24 stores articles (corporate information articles) automatically created by the MPU 20a in accordance with the provisions of the article generation program P2 for each company corresponding to the article. The evaluation criteria table 25 stores conditions for various comparisons used in the analysis and evaluation processing of corporate information.

[0067] FIG. 4 shows an example of the contents of the quarterly financial results table section 25a included in the evaluation criteria table 25. The quarterly financial results table section 25a shown in FIG. 4 is used when new company information is announced as quarterly financial results, and specifies numerical values ​​of points (corresponding to evaluation values ​​common to different items) indicating the evaluation to be assigned as the analysis result. The quarterly financial results table section 25a includes, as examples of analysis and evaluation items, "(1) Comparison of revised earnings forecast and previous earnings forecast when quarterly financial results are revised," "(2) Comparison of revised dividend forecast and previous dividend forecast when quarterly financial results are revised," "(3) Comparison of actual figures and disclosed forecast when quarterly financial results are not revised," "(4) Comparison of financial results with past results," "(5) Comparison of financial results with the highest ever results," and "(6) Comparison of profit margin and past results."

[0068] "(1) Comparison of revised earnings forecasts with previous earnings forecasts when quarterly financial results are revised" applies when the newly announced corporate information is a revision of quarterly financial results (revision of earnings forecasts, etc.), and specifies the number of points to be awarded (or subtracted) depending on the results of comparing the revised earnings forecast with the previous (pre-revision) earnings forecast. Specifically, if the revision rate of the earnings forecast from the previous rate is 15% or more but less than 30% (15% ≦ revision rate < 30%), one point will be awarded (plus one point); if the revision rate of the earnings forecast from the previous rate is 30% or more (30% ≦ revision rate), two points will be awarded (plus two points); if the revision rate of the earnings forecast from the previous rate is minus 15% or less but less than 30% (-30% < revision rate ≦ -15%), one point will be subtracted (minus one point); and if the revision rate of the earnings forecast from the previous rate is minus 30% or less (-30% ≧ revision rate), two points will be subtracted (minus two points).

[0069] "(2) Comparison of revised dividend forecasts with previous dividend forecasts when quarterly financial results are revised" also applies when the announced corporate information includes revisions to quarterly financial results (such as dividend forecasts), and specifies the number of points to be awarded (or subtracted) based on the results of comparing the revised dividend forecast with the previous (pre-revision) dividend forecast. Specifically, if the dividend forecast revision rate is 15% or more but less than 30% (15% ≤ revision rate < 30%), one point will be awarded (plus one point). If the dividend forecast revision rate is 30% or more (30% ≤ revision rate), two points will be awarded (plus two points). On the other hand, if the dividend forecast revision rate is minus 15% or less but less than 30% (-30% < revision rate ≤ -15%), one point will be subtracted (minus one point). If the dividend forecast revision rate is minus 30% or less (-30% ≥ revision rate), two points will be subtracted (minus two points).

[0070] "(3) Comparison of actual figures with disclosed forecasts when quarterly financial results are unrevised" applies when the announced corporate information is quarterly financial results that do not include revisions, and specifies the number of points to be awarded (or subtracted) depending on the progress rate obtained by comparing the announced earnings notification with the disclosed earnings forecast. Specifically, if the progress rate against the disclosed forecast is 80% or more (80% ≦ progress rate), 1 point will be awarded (plus 1 point), while if the progress rate is less than 60%, 1 point will be subtracted (minus 1 point).

[0071] "(4) Comparison of Financial Results with Past Performance" applies when the announced corporate information is quarterly financial results, and specifies the number of points to be awarded (or subtracted) based on the comparison between the announced financial results and past performance (the performance of the most recent fiscal year). Specifically, if the announced financial results are 15% or more but less than 30% higher than past performance (15% higher than past performance ≦ financial results < 30% higher than past performance), one point will be awarded (plus one point); if the announced financial results are 30% or more higher than past performance (30% higher than past performance ≦ financial results), two points will be awarded (plus two points); if the announced financial results are 15% or less but less than 30% lower than past performance (-30% lower than past performance < financial results ≦ -15% higher than past performance), one point will be subtracted (minus one point); and if the announced financial results are 30% or less lower than past performance (-30% higher than past performance ≧ financial results), two points will be subtracted (minus two points).

[0072] "(5) Comparison of financial results with the highest ever performance" also applies when the announced corporate information is quarterly financial results or forecast financial results, and stipulates that points are awarded based on whether the announced financial results have been updated (exceeded) by comparing them with the highest ever performance, or whether the forecast financial results included in the announcement are expected to update the highest ever performance. Specifically, if the announced financial results have updated the highest ever performance, one point is awarded (plus one point), and if the announced forecast financial results are expected to update the highest ever performance, one point is awarded (plus one point).

[0073] "(6) Comparison of profit margins with past performance" applies when the disclosed corporate information is quarterly financial statements, and specifies the value of points to be awarded (or deducted) depending on the results of comparing the profit margins included in the disclosed corporate information with past performance (for example, the profit margins for the most recent period or the profit margins for the same period of the previous year, etc.; the same applies below). Specifically, if the profit margin included in the announced corporate information is 15% or more but less than 30% compared to past performance (15% above past performance ≦ profit margin < 30% above past performance), 1 point will be awarded (plus 1 point); if the announced profit margin is 30% or more compared to past performance (30% above past performance ≦ profit margin), 2 points will be awarded (plus 2 points); on the other hand, if the profit margin included in the announced corporate information is 15% or less but less than 30% compared to past performance (-30% above past performance < profit margin ≦ -15% above past performance), 1 point will be deducted (minus 1 point); and if the announced profit margin is 30% or less compared to past performance (-30% above past performance ≧ profit margin), 2 points will be deducted (minus 2 points).

[0074] Fig. 5 shows an example of the contents of the annual settlement table section 25b included in the evaluation criteria table 25. The annual settlement table section 25b shown in Fig. 5 is used when the corporate information to be analyzed and evaluated is an annual settlement, and specifies the numerical value of points (evaluation value) to be assigned as a result of the analysis and evaluation, etc. The annual settlement table section 25b includes, as items of analysis and evaluation, "(10) Comparison of current period's performance forecast with past actual results," "(11) Comparison of settlement figures with the highest ever actual results," "(12) Comparison of previous period's actual results with forecasts," "(13) Comparison of settlement figures in three-month increments with past actual results," and "(14) Comparison of profit margins with past figures."

[0075] "(10) Comparison of current fiscal year earnings forecast with past performance" specifies the value of points to be awarded based on the results of comparing the current fiscal year earnings forecast (e.g., forecast of operating profit) included in the published corporate information (financial results) with past performance figures (e.g., operating profit for the most recent fiscal year). Specifically, if the current fiscal year's earnings forecast included in the announced corporate information is 30% or more but less than 50% above past actual results (30% above past actual results ≦ current fiscal year's earnings forecast < 50% above past actual results), one point will be awarded (plus one point); if the current fiscal year's earnings forecast is 50% or more above past actual results (50% above past actual results ≦ current fiscal year's earnings forecast), two points will be awarded (plus two points); on the other hand, if the current fiscal year's earnings forecast is 30% or less but less than 50% below past actual results (-50% of past actual results < current fiscal year's earnings forecast ≦ -30% of past actual results), one point will be deducted (minus one point); and if the current fiscal year's earnings forecast is 50% or less below past actual results (-50% of past actual results ≧ current fiscal year's earnings forecast), two points will be deducted (minus two points).

[0076] "(11) Comparison of financial results with the highest ever performance," like the above-mentioned "(5) Comparison of financial results with the highest ever performance," provides for points to be awarded based on whether the financial results included in the announced corporate information (financial results) are updated (exceeded) with the highest ever performance (highest ever performance value), or whether the financial results forecast included in the announced corporate information (financial results forecast) is expected to update the highest ever performance. Specifically, if the financial results included in the announced corporate information update the highest ever performance, one point will be awarded (plus one point), and if the announced financial results forecast is expected to update the highest ever performance, one point will be awarded (plus one point).

[0077] "(12) Comparison of previous period's actual figures with forecasts" specifies the value of points to be awarded based on the results of comparing the previous period's actual figures (e.g., the previous period's operating profit) included in the published corporate information (financial results) with the forecast figures for the next period (e.g., the forecast figures for the next period's operating profit). Specifically, if the previous period's actual value included in the published corporate information is 30% or more but less than 50% above the forecast value for the next period (30% above the forecast value for the next period ≦ previous period's actual value < 50% above the forecast value for the next period), one point will be awarded (plus one point); if the previous period's actual value is 50% or more above the forecast value for the next period (50% above the forecast value for the next period ≦ previous period's actual value), two points will be awarded (plus two points); on the other hand, if the previous period's actual value is 30% or less but less than 50% above the forecast value for the next period (-50% of the forecast value for the next period < previous period's actual value ≦ -30% of the forecast value for the next period), one point will be deducted (minus one point); and if the previous period's actual value is 50% or less below the forecast value for the next period (-50% of the forecast value for the next period ≧ previous period's actual value), two points will be deducted (minus two points).

[0078] "(13) Comparison of financial figures for three months with past performance" specifies the value of points to be awarded according to the results of comparing the financial figures for three months (for example, financial figures for three months) included in the published corporate information (financial results) with the performance for the same period of the previous year (for example, financial results figures). Specifically, if the financial figures for three months included in the published corporate information are 30% or more but less than 50% higher than the performance figures for the same period of the previous year (30% higher than the performance figures for the same quarter of the previous year ≦ financial figures for three months < 50% higher than the performance figures for the same quarter of the previous year), one point will be awarded (plus one point), and if the financial figures for three months are 50% or more higher than the performance figures for the same period of the previous year (50% higher than the performance figures for the same period of the previous year ≦ financial figures for three months), two points will be awarded (plus one point). On the other hand, if the financial figures for three months are minus 30% or less but less than 50% of the actual figures for the same period of the previous year (-50% of the actual figures for the same period of the previous year < financial figures for three months ≦ -30% of the actual figures for the same period of the previous year), one point will be deducted (minus 1 point), and if the financial figures for three months are minus 50% or less of the actual figures for the same period of the previous year (-50% of the actual figures for the same period of the previous year ≧ financial figures for three months), two points will be deducted (minus 2 points).

[0079] "(14) Comparison of profit margins with past results" specifies the value of points to be awarded (or subtracted) depending on the results of comparing the profit margins included in the published corporate information with past results (profit margins for the previous period), just as in the case of "(6) Comparison of profit margins with past performance" mentioned above. Specifically, if the profit margin included in the announced corporate information is 15% or more but less than 30% compared to past performance (15% above past performance ≦ profit margin < 30% above past performance), 1 point will be awarded (plus 1 point); if the announced profit margin is 30% or more compared to past performance (30% above past performance ≦ profit margin), 2 points will be awarded (plus 2 points); on the other hand, if the profit margin included in the announced corporate information is 15% or less but less than 30% compared to past performance (-30% above past performance < profit margin ≦ -15% above past performance), 1 point will be deducted (minus 1 point); and if the announced profit margin is 30% or less compared to past performance (-30% above past performance ≧ profit margin), 2 points will be deducted (minus 2 points).

[0080] Note that the conditions related to points (including the numerical values ​​associated with the conditions) shown in the quarterly financial results table section 25a in Fig. 4 and the full-year financial results table section 25b in Fig. 5 are merely examples, and it is of course possible to apply other conditions, etc. For example, the conditions related to dividends were explained in "(2) Comparison of revised dividend forecast and previous dividend forecast when quarterly financial results are revised" in the quarterly financial results table section 25a in Fig. 4 for the case where quarterly financial results are revised, but even for quarterly financial results and full-year financial results without revisions, when compared with the previous period, it is assumed that the following conditions are also specified: if the newly announced dividend is an increase of 5 yen to less than 10 yen, plus 1 point; if it is an increase of 10 yen or more, plus 2 points; if it is a decrease of 5 yen to less than 10 yen, minus 1 point; and if it is a decrease of 10 yen or more, minus 2 points.

[0081] Furthermore, each of the table sections 25a and 25b shown in FIGS. 4 and 5 is a part of the evaluation criteria table 25 and includes comparison conditions with other targets. For example, in addition to past performance (including the highest or lowest performance ever recorded) and forecast values ​​for the next fiscal year, the evaluation criteria table 25 also includes the average value (average numerical information) of multiple companies for each item as a comparison target. When using this average numerical information as a comparison condition, similar to the case of past performance or forecast values ​​for the next fiscal year, provisions are included for awarding plus or minus 1 or 2 points, etc., depending on the comparison condition with the numerical information included in the published company information. Furthermore, the evaluation criteria table 25 also includes a table section containing content specifying the conditions for points, as shown in FIGS. 4 and 5, even when the published company information includes dividend forecasts or dividend forecast revisions. Next, the article generation program P2 stored in the storage unit 20g will be described in detail.

[0082] 6 is a flowchart illustrating the control processes performed by the MPU 20a of the article generation device 20 as defined by the article generation program P2, and the process steps shown in this flowchart correspond to the article generation method. Below, the process steps performed by the MPU 20a as defined by the article generation program P2 will be described according to the flowchart in FIG.

[0083] First, the article generation device 20 (MPU 20a) monitors the status of the information DB system 10 in accordance with the provisions of the article generation program P2, and detects whether the information DB system 10 (DB management device 11) has received company information distributed from the stock exchange system 2 (S1). As described above, when the information DB system 10 receives company information distributed from the stock exchange system 2, the DB management device 11 transmits an information signal indicating which company's company information was received and when, and the article generation device 20 (MPU 20a) performs the above detection by monitoring whether the information DB system 10 has received the company information based on the presence or absence of this information signal.

[0084] If the information DB system 10 does not detect that it has received company information (S1: NO), the article generation device 20 (MPU 20a) enters a state of waiting for detection. On the other hand, if the information DB system 10 detects that it has received company information (S1: YES), the article generation device 20 (MPU 20a) sends a request for various information about the company related to the company information that was detected to have been received (company information request) to the information DB system 10 (DB management device 11) (S2).

[0085] Then, in response to the transmission of the company information request, the article generation device 20 (MPU 20a) receives and acquires the company information (files of each company information) sent from the information DB system 10 (DB management device 11) (S3). For example, if a company information request about Company A is sent at the above-mentioned stage S2, the article generation device 20 will receive and acquire, as company information, the company information of Company A (new company information announced by Company A whose reception was detected), past company information of Company A, average numerical information of multiple other companies (excluding Company A), performance forecast information of Company A (forecasted company information), theoretical stock price information of Company A, and individual investor forecast stock price information of Company A, etc. from the information DB system 10. Note that the article generation device 20 (MPU 20a) temporarily stores the acquired company information in the RAM 20d or the storage unit 20g for subsequent processing.

[0086] When the article generating device 20 (MPU 20a) acquires the requested company information, it performs calculations, comparisons, and other processes on the acquired company information (numerical information included in the company information) (S4). Specific examples of the contents of the calculations, etc. include, if the acquired company information is based on the announcement of quarterly financial results, calculation of the progress rate of the announced actual figures against the business performance forecast, calculation of the progress rate at the same point in time in the past, etc., and if the announced company information involves a revision, calculation of the revision rate, etc., and further calculation of points defined in the evaluation criteria table 25 (table sections 25a, 25b) shown in Figures 4 and 5 described above.

[0087] The contents of the comparisons include a comparison of announced performance values ​​(financial results, profit margins, etc.) with past performance values ​​(past financial results, profit margins, etc.), a comparison of performance values ​​(financial results, profit margins, etc.) for each item in the newly announced corporate information with forecast performance values ​​(forecast financial results and forecast profit margins) for each item included in the forecast performance information, a comparison of announced forecast performance values ​​with past actual values, a comparison of numerical information for each item in the newly announced corporate information with average numerical information for each item, a comparison of the actual progress rate calculated by the above-mentioned calculation with the forecast progress rate, a comparison of the actual progress rate with the forecast progress rate, and a comparison of the actual progress rate with the past progress rate for the same period. Note that these comparisons involve a comparison of numerical information for the same item.

[0088] Among these comparisons, if the acquired corporate information is for full-year financial results, comparisons are made between the newly announced earnings forecast for the current period and past results, comparisons between the announced actual values ​​for the previous period and forecast values, comparisons between three-month financial figures and past results, comparisons between announced profit margins and past profit margins, etc. In comparisons related to full-year financial results, a comparison is always made between financial figures and past results to see whether a new record high has been achieved or whether a new record high is expected, and in this case, a comparison is made between the extreme values ​​of the past (highest or lowest financial values) from the numerical information (financial figures) included in the past corporate information and the newly announced financial figures.

[0089] Furthermore, if the acquired corporate information is quarterly financial results and full-year forecasts for performance or dividends have been revised, the revised forecast performance will be compared with the previous forecast, and the revised forecast dividend will be compared with the previous forecast, etc. On the other hand, if no revisions have been made, the progress rate of actual figures against the forecast will be compared, the progress rate with financial results for the same period in the past will be compared, the financial figures for three months will be compared with past actual figures, and the profit margin will be compared with past profit margins, etc. Even in such comparisons related to quarterly financial results, when comparing financial figures with past results, a comparison is always made to see whether the financial figures have set a new record high (or a new record low) or whether a new record high (or a new record low) is expected, and in this case too, the past extreme values ​​(the highest or lowest financial figures) from the numerical information (financial figures) contained in the past corporate information will be compared with the newly announced financial figures.

[0090] Furthermore, at step S4, the article generation device 20 (MPU 20a) also analyzes whether the published corporate information contains any revisions, and also analyzes which items in the published corporate information will affect the stock price (referred to as stock price drivers). All financial items included in the published corporate information (sales, net profit, EBITDA (earnings before interest, taxes, depreciation, and amortization), net assets, dividends, etc.) are candidates for stock price drivers. The article generation device 20 (MPU 20a) compares the correlation between the numerical values ​​of these items and stock prices (including stock prices distributed from the stock price distribution system 3) with the average prices of other companies sent from the information DB system 10 to identify stock price drivers. Note that stock price drivers affect stock prices, and are therefore of interest to investors.

[0091] 6, the article generation device 20 (MPU 20a) performs processing to identify the headline or subject of the article based on the processing content of S4 described above. That is, in the processing of S4, the article generation device 20 (MPU 20a) compares the latest published company information (financial results, etc.) from among the received company information with past company information (financial results, etc.) and performance forecast information (forecasts related to financial results, etc.) for each item, and performs processing to identify the headline or subject content of the article to be automatically generated based on the results of this comparison at the step of S5.

[0092] The comparison that is the premise of this identification process is to compare, item by item, multiple numerical values ​​(numerical information) related to the company's performance contained in each piece of information at step S4 of the flowchart, and based on the results of this comparison, the article generation device 20 (MPU 20a) calculates points defined in the evaluation criteria table 25. Then, the article generation device 20 (MPU 20a) identifies the item related to the numerical information for which the situation corresponding to the difference between the numerical information was the largest, identifies the content (information) corresponding to that item as a headline or a main topic, and describes the identified content in the headline or main topic. In this embodiment, the calculated points (corresponding to the evaluation value corresponding to the difference) are used as the situation corresponding to the difference between the numerical information, and the item to be compared for which the calculated point value was the largest is identified, and describes the content (information) corresponding to that item in the headline or main topic.

[0093] As described above, the company information acquired by the article generation device 20 (MPU 20a) from the information DB system 10 is presented in a predetermined standardized file format (e.g., XBRL format) (although the numerical information has already been converted to standard numerical units, as described above). In this standardized file format, multiple items representing business performance are listed in a predetermined order, with items and numerical values ​​paired together. Therefore, if the announced company information is financial results, the numerical values ​​indicated by the numerical information for each item included in the newly announced financial results are compared with the numerical values ​​indicated by the numerical information for each item included in the past financial results and the numerical values ​​indicated by the numerical information for each item included in the forecast financial results. Items included in the financial results include sales, operating profit, profit before tax, ordinary profit, total assets, dividends, etc., and the numerical values ​​for each of these items are compared sequentially.

[0094] Furthermore, for the comparison results, points (evaluation values) are calculated based on the specifications in the evaluation criteria table 25 (each table section 25a, 25b) shown in Figures 4 and 5 described above, and the article generation device 20 (MPU 20a) identifies the item to be compared that has the largest absolute value of the points (corresponding to the difference situation), and decides to use information corresponding to the identified item as the headline or method of the article (S5). In this way, in this embodiment, the headline or method is decided by identifying the item with the largest absolute value of the points (evaluation values), so that not only when the operating profit of the latest announced financial statements is positive compared to the operating profit of past financial statements, but also when it is negative, if the evaluation value corresponding to the difference (difference in absolute values) is the largest among multiple items, a statement such as "operating profits have dropped significantly" will be used as the headline or subject.

[0095] Note that using absolute values ​​as the situation according to the difference (points, which are the evaluation values) is just one example, and depending on the system specifications, it may be possible to specify items only when the evaluation value (points) according to the difference is positive, without using absolute values, which is preferable when it is desired to convey to the user the company's most positive item in an easy-to-understand manner.On the other hand, it is also possible to specify items only when the evaluation value (points) according to the difference is negative, which is preferable when it is desired to highlight and convey to the user the company's most negative item (item with the largest negative value).

[0096] Then, at step S6 of the flowchart shown in Figure 6, the article generation device 20 (MPU 20a) performs a process for determining the content and order of the paragraphs that make up the article. In identifying the content of the paragraphs, since operating profit is basically the most important criterion (important item) among the multiple items included in the company information, the operating profit item is always included in the content of the paragraph. Also, in cases where the operating profit status and the other profit status are opposite, such as when the published company information shows an operating profit in the black but a net profit in the red, items other than operating profit can be factors that affect stock prices, so such items that can affect stock prices are treated as important items, and it is determined that information related to those important items will also be included in the content of the paragraph.

[0097] Furthermore, the processing related to stock price drivers described above is also used in the processing for determining the content and order of paragraphs. That is, numerical values ​​(numerical information) corresponding to multiple items contained in the acquired new company information are compared with the average numerical information contained in the information acquired from the information DB system 10, and based on the comparison results, points indicating the situation corresponding to the difference between the two numerical information (proportion corresponding to the difference: equivalent to the evaluation value of the difference) are calculated in accordance with the evaluation criteria table 25. The order of items in which the points of the items related to these calculated points increase is identified, and information related to each item is entered in accordance with the identified order of items to automatically generate an article. In this case, the description of the information related to each item is divided into paragraphs for each description content, thereby determining the content and order of the paragraphs.

[0098] 6, the article generation device 20 (MPU 20a) automatically generates an article about company information based on the details of the processes in S4, S5, S6, etc. Specifically, for the headline of the article, the article generation device 20 (MPU 20a) creates a headline sentence describing information corresponding to the item identified in S5 using a function (AI function) corresponding to the sentence creation engine included in the article generation program P2.

[0099] Regarding the content of the article, the article generation device 20 (MPU 20a) uses functions corresponding to the writing engine to create the first paragraph, which includes the subject identified in step S5 and the important items determined in step S6, and also creates the second and subsequent paragraphs by including information corresponding to the items determined in step S6 in each paragraph.

[0100] In creating such sentences, since the XBRL formatted corporate information obtained from the information DB system 10 is a set of the above-mentioned items and numerical information, the items are basically recognized as subjects within these sets, and the numerical information is recognized as objects. The particles, verbs, etc. connecting the subject items and the object numerical information are identified through analysis by the sentence creation engine, and specific sentences are created for each paragraph. In other words, the words and numerical values, etc. corresponding to the subject and predicate used in each paragraph are created using the item contents and numerical information corresponding to the operating profit, important items, etc., processed in the above-mentioned step S6, etc., as well as the information on the performance forecast included in the corporate information obtained in step S3 (information stored in the performance forecast DB 17) and the contents of the results calculated by the calculation in step S4 (calculated items and calculated numerical values), etc.

[0101] For the first sentence of the first paragraph, the company name, the date and time of the announcement, and the type of announcement (financial results, earnings forecast revision, dividend revision, etc.) contained in the newly released company information are extracted, and the sentence creation engine analyzes these extracted words to determine particles, adverbs, verbs, etc. that connect them appropriately to create the first sentence of the first paragraph. The company information articles created in this way are then stored by the MPU 20a in the created article table 24 (see Figure 3) contained in the storage unit 20g, separated by company corresponding to the article.

[0102] Finally, at step S8 in the flowchart shown in FIG. 6, the article generation device 20 (MPU 20a) distributes the article (corporate information article) generated at step S7 above to user terminals T1, T2, T3, etc. shown in FIG. 1 via a network. This distribution can be sent to the target user terminal in a push-distribution manner, or to a requesting user terminal in response to a request from the user terminal. In commercial use, a fee may be charged for the distribution. Furthermore, the distribution of the corporate information article at step S8 from step S1 can occur as quickly as instantaneously (up to 10 seconds) or within one minute, and typically occurs within two to three minutes. Even in cases where multiple announcements are made at the same time and multiple articles are generated in parallel, the distribution can occur within four to six minutes at the latest.

[0103] Fig. 7 shows a company information article 30, which is an example of an article automatically generated by the processing of the flowchart shown in Fig. 6. This company information article 30, like article K10 shown in Fig. 14(a), was automatically generated by the present invention based on the financial statements (company information: financial results) of Company A disclosed (announced) by a stock exchange system 2 such as the Tokyo Stock Exchange.

[0104] The company information article 30 has a headline section 30a, a first paragraph section 30b, a second paragraph section 30c, a third paragraph section 30d, and a distribution time section 30e. The headline section 30a indicates that Company A's operating profit for the May-January period increased by 51% and that it also increased by 2.3 times for the November-January period. These operating profit-related details were identified in step S5 of the flowchart in Figure 6. As shown in Figure 7, the content of this headline section 30a succinctly indicates the operating profit status of Company A, and differs from the uniform headline content of article K10 based on the same company A's prior art, shown in Figure 14(a).

[0105] The first paragraph 30b of the company information article 30 begins with a sentence indicating the company name (Company A), the date and time of the announcement (March 1, 3:00 PM), and the type of announcement (financial results). It then goes on to include the main and important items, such as consolidated operating profit for the third quarter compared to the same period last year (which increased 51.4% to ¥17.5 billion), progress against the full-year plan (forecast), and a comparison with the past five years. This first paragraph 30b clearly communicates the operating profit situation to readers and also indicates progress against the plan, accurately conveying information about the company that investors are closely watching. The progress rate and comparison with the five-year average in the first paragraph 30b are based on the processing performed at step S4 in the flowchart shown in Figure 6. On the other hand, in article K10 based on the conventional technology in Figure 14(a), the first paragraph lists the contents of four items: net profit, sales, operating profit, and operating profit, making it difficult for readers of the article to understand which items are important.Therefore, the content regarding operating profit, which was written prominently in article information article 30 of the present invention, is buried in the list of items in article K10.

[0106] Furthermore, the second paragraph 30c of the company information article 30 compares the forecast for consolidated operating profit for the February-April period with the same period last year, providing a forecast of the company's situation for the next fiscal year, which investors are concerned about. This forecast for the next fiscal year is based on a comparison of the previous fiscal year's actual results and the forecast for the next fiscal year, as specified in the evaluation criteria table 25 (step S4 in the flowchart of Figure 6). Meanwhile, in the article K10 based on the prior art shown in Figure 14(a), the forecast for the next fiscal year is stated in the second paragraph. However, like the first paragraph, the forecast for the four items of net profit, sales, operating profit, and operating profit is listed in a list format, making it difficult for readers to understand which items are important.

[0107] Furthermore, the third paragraph 30d of the company information article 30 compares the consolidated operating profit and operating profit margin, which are the results for the most recent three months, with the same period last year, and the content of this third paragraph 30d is useful information for investors to understand the most recent business performance of Company A. On the other hand, the prior art article K10 in Figure 14(a) does not include content that corresponds to the third paragraph 30d of the company information article 30.

[0108] To summarize the above, if investors or the like read the company information article 30 about Company A shown in Figure 7 according to the present invention, the contents of the headline section 30a and the first paragraph section 30b will show that Company A's current business performance is going well, which is also supported by the contents of the third paragraph section 30c, but the contents of the second paragraph section 30c will clearly convey the information that allows one to determine that the current business performance is at its peak, as business performance will tend to decline in the future.However, it can be seen that it is difficult to read the above content, which can be determined from the company information article 30 in Figure 7, from article K10 based on the prior art in Figure 14(a).

[0109] Furthermore, Figure 8 shows a corporate information article 31, which is another example of an article automatically generated by the processing of the flowchart shown in Figure 6. This corporate information article 31, like article K11 shown in Figure 14(b), was automatically generated based on the financial statements (corporate information: financial statements) of Company B disclosed (announced) by a stock exchange system 2 such as the Tokyo Stock Exchange.

[0110] Like the company information article 30 shown in Figure 7, the company information article 31 has a headline section 31a, a first paragraph section 31b, a second paragraph section 31c, a third paragraph section 31d, and a distribution time section 31e. The headline section 31a indicates that Company B's current operating profit will increase by 3%, its highest profit in five consecutive years, and that its dividend will increase by 11 yen. The content of this headline section 31a was identified in step S5 of the flowchart shown in Figure 6. The content of this headline section 31a is different from the content of the headline of article K11 based on the same company B's prior art shown in Figure 14(b), which is about net profit (Company B's net profit for the fiscal year ending January 2017 was 121.5 billion yen, a 44.5% increase). Therefore, the headline content of the company information article 31 about Company B is different, just like the articles shown in Figures 7 and 14(a).

[0111] The first paragraph 31b of this company information article 31 includes the company name (Company B), the date and time of the announcement (March 9, 3:30 PM), and the type of announcement (financial results), followed by a main or important item: a comparison of consolidated operating profit for the fiscal year ending January 2017 with the previous fiscal year (190.9 billion yen, an 18.9% increase from the previous fiscal year), and an indication that operating profit for the fiscal year ending January 2018 is also expected to increase compared to the previous fiscal year, setting a new record high for the fifth consecutive fiscal year. This makes it easy for readers to understand that Company B is performing well and that this trend is likely to continue. The content of the first paragraph 31b (current status, future details, record high operating profit, etc.) was obtained by the processing in steps S4 to S6 of the flowchart shown in Figure 6.

[0112] The content of the first paragraph 31b of this company information article 31 is different from the content of the first paragraph 30b of the performance information article 30 shown in Figure 7 above in that it includes future performance forecasts, and therefore it can be seen that in the present invention, articles with individual content are automatically generated according to the performance of each company. Also, in article K11 using the prior art shown in Figure 14(b), which is also an article about Company B, the first paragraph is limited to a formulaic listing of the contents of four items - net profit, sales, ordinary profit, and operating profit - and, like article K11 in Figure 14(a), the expression makes it difficult for readers of the article to understand which items are important.

[0113] Furthermore, the second paragraph 31c of the company information article 31 details an increase in the company's annual dividend for the current fiscal year, conveying the dividend information that investors are concerned about. This dividend information was obtained through processing at steps such as S4 in the flowchart of FIG. 6. The second paragraph 31c of the company information article 31 differs from the second paragraph 30c of the company information article 30 shown in FIG. 7, demonstrating that individual article content is generated for each company. Furthermore, in the article K11 based on the prior art shown in FIG. 14(b), which is also about Company B, the second paragraph includes a forecast for the next fiscal year, resulting in completely different content from the second paragraph 31c of the company information article 31.

[0114] Furthermore, the third paragraph 31d of the company information article 31 compares the consolidated operating profit and operating profit margin for the most recent three months with the same period of the previous year, and the content of this third paragraph 31d is useful information for investors to understand the most recent business performance information of Company B. On the other hand, the prior art article K11 of Figure 14(b) does not include content corresponding to the third paragraph 31d of the company information article 31.

[0115] To summarize the above regarding the company information article 31 in Figure 8, when investors and the like read the company information article 31 about Company B shown in Figure 8 according to the present invention, the contents of the headline section 31a and the first to third paragraph sections 31b to 31d clearly convey the information that allows them to determine that Company A's current performance is good and that good performance is likely to continue for some time to come.However, from article K11 based on the conventional technology in Figure 14(b), it is clear that it is difficult to smoothly grasp the above content that can be read from the company information article 31 in Figure 8.

[0116] Figure 9(a) shows a corporate information article 40, which is an example of an article automatically generated by the processing of the flowchart shown in Figure 6. This corporate information article 40 was automatically generated based on materials (corporate information: earnings forecast revisions) from Company C that were disclosed (announced) by a stock exchange system 2 such as the Tokyo Stock Exchange.

[0117] The company information article 40 includes a headline section 40a, a first paragraph section 40b, a second paragraph section 40c, and a distribution time section 40d. The headline section 40a indicates that Company C has revised its current fiscal year operating profit upward by 16%, which was identified in step S5 of the flowchart in Figure 6. The first paragraph section 40b of the company information article 40 includes, as main or important items, the upward revision of the company's consolidated operating profit for the fiscal year ending March 2017 from the previous forecast, indicating that the rate of decline is expected to decrease, and the downward revision of the company's consolidated net profit for the full fiscal year from the previous forecast, indicating that the rate of decline is also expected to increase. This accurately conveys the company's performance trends, which are of interest to investors. The content of the first paragraph section 40b is also based on the processing performed in steps S4 to S6 of the flowchart in Figure 6.

[0118] Furthermore, second paragraph 40c of company information article 40 indicates that the announcement of this revised forecast will result in an increase in consolidated ordinary profit and profit growth rate for the second half of the year, and this content is also based on the processing content at stages S4 to S6 of the flowchart shown in Fig. 6. In this way, when a company announces a revised earnings forecast, the present invention automatically generates an article informing investors of content that will affect stock prices, whereas Non-Patent Document 1 of the prior art mentioned above generates articles only when a company releases financial statements, and therefore cannot generate articles when a revised earnings forecast is announced like this.

[0119] Figure 9(b) also shows an example of a company information article 41 that was automatically generated by processing the flowchart shown in Figure 6, and this company information article 41 was automatically generated based on Company D's materials (company information: earnings forecast revisions) disclosed (announced) by a stock exchange system 2 such as the Tokyo Stock Exchange.

[0120] The company information article 41 also has a headline section 41a, a first paragraph section 41b, a second paragraph section 41c, and a distribution time section 41d. The headline section 41a indicates that Company D has revised its current fiscal year operating profit upward by 16%, setting a new record for profits. This content is also identified in steps S5 and other sections of the flowchart in Figure 6. The first paragraph section 41b of the company information article 41 contains, as a main or important item, a statement that the company has revised its consolidated operating profit for the fiscal year ending March 2017 upward from its previous forecast, now expecting a 16% increase, and is now forecasting a new record for profits for the second consecutive fiscal year. The second paragraph section 41c indicates that, as a result of this revised forecast, the company's consolidated operating profit and profit growth rate for the second half of the fiscal year will increase and expand, respectively. This content is also obtained by processing steps S4 to S6 of the flowchart in Figure 6.

[0121] Figure 10 shows a corporate information article 42, which is an example of an article automatically generated by the processing of the flowchart shown in Figure 6. This corporate information article 42 was also automatically generated based on the materials (corporate information: dividend revision) of Company E disclosed (announced) by a stock exchange system 2 such as the Tokyo Stock Exchange.

[0122] The company information article 42 has a headline section 42a, a first paragraph section 42b, and a distribution time section 42c. The headline section 42a states that Company E will resume its current dividend of 20 yen for the first time in two years. This content is identified in steps S5 and other sections of the flowchart in Figure 6. Furthermore, the first paragraph section 42b of the company information article 42 indicates that the previously undetermined dividend for the fiscal year ending March 2017 will be paid, resuming dividends for the first time in two years, thereby accurately conveying the company's status of concern to investors. The content of this first paragraph section 42b is also based on the processing content in steps S4 to S6 of the flowchart in Figure 6. Thus, in this invention, an article is automatically generated even when a company announces a dividend adjustment. Meanwhile, in the aforementioned Non-Patent Document 1, articles are generated only when a company releases financial statements, and therefore an article cannot be generated when a dividend adjustment is announced.

[0123] As described above, with this invention, when each company announces corporate information such as financial results, an article that focuses on noteworthy matters in accordance with the content of the announcement, rather than a standardized content, is automatically created. This has the advantage of being able to quickly provide investors with articles that use easy-to-understand language about matters that must be considered and judged when making investments, etc. Another feature is that articles with such content are also generated when announcements other than financial results, such as earnings forecast revisions and dividend revisions, are announced.

[0124] The present invention is not limited to the above-described embodiment, and various modifications are possible. For example, in the above description, the average value stored in the past settlement DB 16 is calculated by excluding one company from the remaining companies. However, the other companies may be limited to companies in the same industry as the one company, and an average value may be calculated for each industry, and comparisons may be performed according to the industry. Furthermore, in order to simplify the specifications and reduce the burden of the average calculation process, multiple average values ​​including the one company may be calculated, rather than an average value excluding one company, and the average values ​​calculated in this way may be used for comparisons, etc.

[0125] Furthermore, although the article to be generated includes a heading section 30a or the like indicating the article's headline, depending on the specifications of the article to be generated, the heading section 30a or the like may be omitted. In this case, the content identified at step S5 of the flowchart in Figure 6 is written in the subject of the article, and the subject of the article is included in the sentence following the sentence about the company name and the publication time in the first paragraph section 30b or the like, making the subject stand out.

[0126] Furthermore, in the above explanation, the headline or subject of the article is identified by comparing the numerical information contained in the newly released company information with both the numerical information contained in the performance forecast information and the numerical information contained in the past company information, but if the specifications are to be simplified, the identification may be made by comparing only one of them (it is also possible to identify the headline, etc. by comparing the numerical information contained in the new company information with the numerical information contained in the performance forecast information, or by comparing the numerical information contained in the new company information with the numerical information contained in the past company information).

[0127] Furthermore, although the articles generated in the above process did not include content related to stock prices, the information DB system 10 also sends information related to stock prices (theoretical stock prices, forecast stock prices by multiple individual investors, etc.) stored in the theoretical stock price DB 18 and the individual investor stock price forecast DB 19 to the article generation device 20, and the article generation device 20 also receives and acquires the latest stock prices (actual stock prices) of each company distributed by the stock price distribution system 3, so this information related to stock prices may be included in the articles to be generated.

[0128] 11(a) shows an example of Company A stock price information 50 including Company A's latest stock price 50a (actual stock price) and Company A's theoretical stock price 50b. Such Company A stock price information 50 may be placed within the article, for example, below the headline section 30a or above the distribution time section 30e of the company information article 30 shown in FIG. 7, so that content related to stock prices is provided together with the article. In this case, the latest stock price obtained from the stock price distribution system 3 when the article is generated is used as Company A's stock price 50a included in Company A stock price information 50, and the theoretical stock price of Company A read from the theoretical stock price DB 18 included in the information obtained at step S3 of the flowchart in FIG. 6 is used as Company A's theoretical stock price.

[0129] 11(b) shows an example of Company A stock price information 51 including Company A's stock price 51a (actual stock price) and Company A's forecast stock price 51b. As in the case of Company A's stock price information 50 in Fig. 11(a), such Company A's stock price information 51 may be placed within the article, for example, below the headline section 30a or above the distribution time section 30e of the company information article 30 shown in Fig. 7, and content related to the stock price may be provided together with the article. In this case, as in the case shown in Fig. 11(a), Company A's latest stock price is used as Company A's stock price 51a included in Company A's stock price information 51, and the forecast stock price of Company A read from Individual Investor Stock Price Forecast DB 19 included in the information acquired at step S3 of the flowchart in Fig. 6 is used as Company A's stock price 51a.

[0130] As another variation, it is also possible to use stock price information that combines Figures 11(a) and (b), in which case stock price information including, for example, the stock price of Company A (latest stock price), the theoretical stock price of Company A, and the forecast stock price of Company A would be posted together with the article. By posting stock price information such as that shown in Figures 11(a) and (b) together with the article, it becomes possible to simultaneously check the impact and correlation between the stock price and the content of the article.

[0131] Furthermore, in automatically generated articles about financial statements (such as corporate information articles 30, 3140-42 shown in Figures 7-10, etc.), an overall evaluation of the announcement (an overall evaluation of the announcement about financial statements, etc.) may be displayed along with the article, so that the outline of the announcement that caused the article to be generated can be seen at a glance.

[0132] FIG. 11(c) shows an example of an overall evaluation 52 (an example of an overall evaluation) for Company A's announced financial results. In this example, five stars are displayed so that the financial results can be evaluated on a five-point scale, and the stars are filled in according to the evaluation (for example, if the evaluation is a three out of five, the three stars from the left are filled in). This five-point evaluation is performed by adding up the points for each item calculated based on the evaluation criteria table 25 and performing the five-point evaluation based on the total points. This evaluation is performed by calculating the points in step S4 of the flowchart in FIG. 6, calculating the total points, and determining the evaluation level based on the total points. For example, if the calculated total points is five, the evaluation is also five out of five. Similarly, if the total points is four, the evaluation is four out of five; if the total points is three, the evaluation is three out of five; if the total points is two, the evaluation is two out of five; and if the total points is one, the evaluation is one out of five.

[0133] The overall rating 52 of Company A generated based on the determined rating is placed, for example, below the headline section 30a or above the distribution time section 30e of the company information article 30 shown in Figure 7, and the rating of the announced financial results is provided together with the article. By providing the overall rating 52 together with the article in this way, investors and the like can understand at a glance from the overall rating 52 whether the announced financial results are good or bad, and can then check the article for more detailed information, which contributes to investors and the like being able to make smooth decisions.

[0134] As for the method of evaluation, in addition to the five-point evaluation mentioned above, it is also possible to increase the number of points to six or more, or conversely, reduce the number of points to three or less. Furthermore, it is also possible to evaluate the overall evaluation in two points, plus or minus.

[0135] Even when making an overall evaluation of the released corporate information in this way using two levels of plus or minus, the numerical information for each item contained in the new corporate information is compared with the numerical information for each item contained in the comparable earnings forecast information or the numerical information for each item contained in the past corporate information, and the results of the comparison are based on points calculated for each item according to the conditions specified in evaluation criteria table 25, as described above. However, if the calculated points for multiple items contain a mixture of pluses and minuses, that is, if the numerical information for each item contained in the newly released corporate information is sometimes higher and sometimes lower than the comparable numerical information, it becomes difficult to make a uniform overall evaluation, and in such cases the overall evaluation is made based on the most important item among the multiple items.

[0136] That is, when comparing the numerical information in the important items, if the numerical information included in the new company information is higher than the numerical information being compared (if the calculated points in the important items are positive), the new company information will be given a positive overall evaluation (comprehensive evaluation). On the other hand, if the numerical information included in the new company information is lower than the numerical information being compared (if the calculated points in the important items are negative), the new company information will be given a negative overall evaluation (comprehensive evaluation). Then, by attaching such a positive or negative overall evaluation to an automatically generated article (such as company information article 30) and distributing it to users such as investors, investors can quickly determine whether the announced company information is generally positive or negative based on the overall evaluation attached to the article.

[0137] Among the multiple items, operating profit is generally used as the important item (items related to operating profit are designated as important items). However, when profit-related items include a mix of profits and losses, such as when operating profit is in the black but net profit is in the red, it can be assumed that items other than operating profit have an impact on the company's performance. In such cases, items other than operating profit that have an impact on performance are designated as important items. To identify specific important items in such cases (when identifying items other than operating profit as important items), the article generation device 20 compares the numerical information corresponding to the multiple items included in the newly released company information described above with the average numerical information for each item, and designates the item to be compared that has the largest value (point) indicating the situation corresponding to the difference between the numerical information and the average numerical information as the important item.

[0138] Furthermore, in the above explanation, the generated articles (such as the company information articles 30, 31, 40-42 shown in FIGS. 7-10) are text-based, but it is also possible to visualize the article contents using graphs, images, and the like.

[0139] FIG. 12 shows a visualized example of a corporate information article 60 (an article about announced financial results), which includes, from top to bottom, a summary section 61, a graph section 62, and an evaluation section 63. In this variation, to visualize the article, the financial results are limited to the items "sales," "operating profit," and "net profit." The summary section 61 includes a summary and a graphic indicating a rating of "Good!" or "Bad" for each of these three items. The graph section 62 represents these three items as a graph with each period on the horizontal axis. The evaluation section 63 is essentially the same as that shown in FIG. 11(c). This visualized corporate information article 60 allows for a more visual understanding of the content than a text article, making it ideal for investors and others who want to quickly understand corporate information. [Industrial Applicability]

[0140] The present invention automatically generates articles that are edited to focus on important matters contained in corporate information such as announced financial results, and is therefore suitable for applications such as providing matters of concern to investors in easy-to-understand language. [Explanation of symbols]

[0141] 1. Article provision system 2. Stock Exchange System 3 Stock price distribution system 5. Article generation system 10 Information DB System 11 DB management device 15 Applicable settlement DB 16 Past financial statements DB 17 Earnings Forecast DB 18 Theoretical Stock Price DB 19 Individual investor stock price forecast database 20 Article Generation Device 20a MPU 20g storage section 24 Created Article Table 25 Evaluation Criteria Table P2 Article Generator

Claims

1. An article generation device that includes a comparison means for comparing numerical information included in company information relating to the business performance of a company distributed from an information distribution source with numerical information to be compared, and that automatically generates a company information article based on the comparison made by the comparison means, The article generation device a level determination means for determining an evaluation level from among a plurality of levels according to a result of the comparison between the numerical information items performed by the comparison means; evaluation generating means for generating a comprehensive evaluation representing the evaluation stage determined by the stage determining means as an overall evaluation of the company information; article generation means for automatically generating a company information article in which the comprehensive evaluation generated by the evaluation generation means is arranged; An article generation device comprising:

2. The article generating device according to claim 1 , wherein the evaluation generating means generates a comprehensive evaluation including a plurality of designs corresponding to the plurality of levels.

3. The article generation device according to claim 2, wherein the evaluation generation means generates an overall evaluation that represents the evaluation stage determined by the stage determination means by filling in a number of patterns among the plurality of patterns corresponding to the evaluation stage determined by the stage determination means.

4. The article generation device according to any one of claims 1 to 3, wherein the article generation means automatically generates a company information article including a summary section that summarizes the article and an evaluation section, and places the overall evaluation in the evaluation section.

5. 5. The article generation device according to claim 1, wherein the article generation means automatically generates a company information article including a graph section that graphically represents numerical information included in the company information and an evaluation section, and places the overall evaluation in the evaluation section.

6. 6. The article generating device according to claim 1, wherein the article generating means automatically generates a company information article in which the overall evaluation is arranged together with text related to the article.

7. The company information is received by an external information database system; a detection means for detecting receipt of company information in an external database system; An article generation device as described in any one of claims 1 to 6, wherein when the detection means detects receipt of company information in an external database system, the comparison means compares the numerical information contained in the company information whose reception has been detected with the numerical information to be compared.

8. The company information includes numerical information corresponding to a plurality of items, the comparison target includes numerical information corresponding to the same plurality of items as the plurality of items of the company information, The comparison means compares the numerical information for each of the same items, The article generation device according to any one of claims 1 to 7, wherein the level determination means calculates an evaluation value for each item according to the result of the comparison between the numerical information for each of the same items performed by the comparison means, and determines the evaluation level by summing up the calculated evaluation values ​​for each item.

9. An article generation method in which an article generation device executes a comparison step of comparing numerical information included in company information relating to company performance distributed from an information distribution source with numerical information to be compared, and automatically generates a company information article based on the comparison in the comparison step, The article generation device a stage determination step of determining an evaluation stage from among a plurality of stages according to a result of the comparison between the numerical information in the comparison step; an evaluation generating step of generating a comprehensive evaluation representing the evaluation stage determined in the stage determining step as an overall evaluation of the company information; a step of automatically generating a company information article in which the comprehensive evaluation generated in the evaluation generating step is arranged; An article generation method comprising:

10. A computer program for causing a computer to execute a comparison step of comparing numerical information included in company information relating to company performance distributed from an information distribution source with numerical information to be compared, and for causing a computer to perform processing for automatically generating company information articles based on the comparison in the comparison step, The computer, a stage determination step of determining an evaluation stage from among a plurality of stages according to a result of the comparison between the numerical information in the comparison step; an evaluation generating step of generating a comprehensive evaluation representing the evaluation stage determined in the stage determining step as an overall evaluation of the company information; a step of automatically generating a company information article in which the comprehensive evaluation generated in the evaluation generating step is arranged; A computer program characterized by causing the computer to execute the following:

Images