JP6767953B2 - Article generation system, article generation device, article generation method, and computer program - Google Patents

Description

The present invention automatically generates articles, and in particular, by comparing the content of announcements regarding the performance of each company with performance forecasts, etc., items that are highly relevant to performance trends are identified, and based on the determined items. It relates to an article generation system, an article generation device, an article generation method, and a computer program that automatically generate articles in which the performance status of each company is individually edited.

Conventionally, sentences such as articles have been automatically generated. Patent Document 1 below discloses generating a textual report with numbers or data values inserted using an appropriate template based on the type of event that occurs (paragraph 0023 of Patent Document 1). ~ 0029, and FIG. 1B, etc.). As an example, a report on the fourth quarter earnings of a company is presented (see paragraphs 0031, 0032, and FIG. 4 of Patent Document 1).

Further, Patent Document 2 below discloses that a commentary on a stock issue designated by a user is generated from stock price data, contents obtained by analyzing stock price data, and news data (paragraph of Patent Document 2). 0022 to 0053, and FIGS. 1, 3, etc.). According to the disclosure contents of Patent Document 2, examples of the explanatory text generated in Patent Document 2 are the explanatory texts K1 and K2 shown in FIGS. 13A and 13B.

Furthermore, in recent years, articles are automatically created by artificial intelligence (AI). For example, in Non-Patent Document 1 below, when a listed company discloses (announces) financial statements on "TDnet (Timely Disclosure Information Browsing Service)" operated by the Tokyo Stock Exchange, artificial intelligence (AI) is based on the financial statements. ) Is summarized in three paragraphs, "performance", "factor", and "outlook", and the published figures are applied to the template to create a sentence, and the reason why sales and profits have changed from the same period of the previous year is also included in the article. The description is disclosed. Examples of articles created by Non-Patent Document 1 are shown in FIGS. 14 (a) and 14 (b). FIG. 14A is an article K10 of a listed company A, and FIG. 14B is an article K11 of a listed company B.

As another example of article creation by artificial intelligence (AI), in Non-Patent Document 2 below, in addition to the above-mentioned "TDnet", it is included in the information of "EDINET" which is an electronic disclosure system regarding disclosure documents of the Financial Services Agency. It is disclosed that the points are provided as news (QUICK AI bulletin).

Special Table 2004-526264 Japanese Unexamined Patent Publication No. 2001-351111 IT Pro- "Making a manuscript in just 10 seconds, the impact of Nikkei's AI reporter" Financial Statement Summary "3D Motion Human Anatomy The world's first 3D human anatomy app that reproduces the movement and morphology of living humans", [online], [Heisei Searched on July 3, 2017], Internet <URL: http://itpro.nikkeibp.co.jp/atcl/column/17/040400118/041200003/> @ Press- "Automatic analysis news" QUICK AI bulletin "release-instantly detect stock price impact-", [online], [Search on July 3, 2017], Internet <URL: https: // www.atpress.ne.jp/news/116363 ＞

Since the report generated by Patent Document 1 has numbers or data values inserted in the template, the content of the report becomes uniform, and what are the important items that affect the stock price, etc. There is a problem that it is difficult to grasp and the overall evaluation is also difficult to understand at a glance. Similarly, the commentary generated by Patent Document 2 is also a standard sentence that describes the rise or fall of the stock price for the stock issue specified by the user, so that important items related to the corporate performance are described. There is a problem that it is difficult to grasp, and there is also a problem that the content is limited to the explanation about stock prices.

Furthermore, the article created by Non-Patent Document 1 is basically a standard sentence in which the numerical values included in the information disclosed in "TDnet" are applied to a template, and FIG. 14 (a) As shown in articles K10 and K11 of (b), there are only a list of multiple items and numerical values related to the items, and what are the important items for investors to make investment decisions? There is a problem that it is difficult to understand at a glance what the overall evaluation is like in the created article.

Furthermore, the news created by Non-Patent Document 2 is basically limited to the content using the information disclosed in "TDnet" and is not compared with the earnings forecast of the company, so it is uniform. There is a problem that it is the content.

The present invention has been made in view of such circumstances, and by comparing the numerical information included in the announcement of the financial statements, dividends, revisions, etc. of each company with the business performance forecast, etc., the business performance, stock price, etc. of the company can be obtained. Article generation system, article generation device, article that automatically generates articles that make it easier for investors to convey the items that investors are paying attention to by identifying the items that affect them and applying such items to the headline or subject of the article. It is an object of the present invention to provide a generation method and a computer program.

Further, in the present invention, items related to business performance, etc. included in the announced content are specified for each presentation, and articles are edited focusing on the specified items, and individually according to the business performance, etc. of each company. It is an object of the present invention to provide an article generation system, an article generation device, an article generation method, and a computer program that automatically generate articles with various contents.

Further, the present invention presents an overall evaluation of a company's announced performance, etc., thereby helping investors to quickly judge the announced performance, etc. Article generation system, article It is an object of the present invention to provide a generator, an article generation method, and a computer program.

In order to solve the above problems, the present invention is an article generation system including an article generation device that automatically generates a company information article based on the company information distributed from the information distribution source, and the company information distributed from the information distribution source. Includes numerical information corresponding to a plurality of items related to the business performance of the company, receives the company information distributed from the information distribution source, and stores the forecasted company information including the numerical information of the forecast according to the plurality of items. The article generator is provided with a means for detecting that new company information distributed from an information distribution source is received by the information database system, and new company information by the information database system. When the reception of the new company information is detected, the means for acquiring the new company information and the predicted company information of the same company as the company related to the new company information from the information database system, and a plurality of information included in the acquired new company information. A means to compare the numerical information according to the item with the numerical information of the forecast according to multiple items included in the acquired forecast company information for each item, and according to the difference between the numerical information in the compared numerical information. It is characterized by including a means for specifying an item related to numerical information having the largest value, and a means for describing information corresponding to the specified item in the heading or subject of a company information article.

Further, in the present invention, the information database system accumulates company information received from an information distribution source as past company information, and the article generator is a new company information obtained by the information database system. When reception is detected, in addition to the new company information and the forecast company information, a means for acquiring the past company information of the same company as the company related to the new company information from the information database system and the new Numerical information according to a plurality of items included in the company information, numerical information of the forecast according to the plurality of items included in the forecast company information, and past according to a plurality of items included in the acquired past company information. It is characterized in that it is provided with numerical information of the above and a means for comparing each item.

Further, in the present invention, the information database system stores average numerical information obtained by averaging numerical information corresponding to a plurality of items included in corporate information of a plurality of companies, and the article generator is a new company. Compare with the means to acquire average numerical information according to the acquisition of information and the means to compare the acquired average numerical information and the numerical information corresponding to multiple items included in the acquired new company information for each item. A means for specifying the order of items having a large value according to the difference between the numerical information and the average numerical information, and a means for automatically generating a company information article by describing information related to each item according to the order of the specified items. It is characterized by having.

Furthermore, in the present invention, the information database system includes means for storing a theoretical stock price theoretically calculated for each company's stock price, and the article generator is said to be in accordance with the acquisition of new company information. Means for acquiring the theoretical stock price of a company according to new company information, means for acquiring the actual stock price of each company to be distributed, the acquired theoretical stock price, and the actual stock price of the same company as the company related to the theoretical stock price. It is characterized by providing a means for automatically generating corporate information articles including stock prices.

Then, in the present invention, the information database system includes means for storing the investor's expected stock price calculated by aggregating the stock prices predicted by a plurality of investors for the stock price of each company, and the article generator is A means of acquiring the investor's expected stock price of a company according to the new company information, a means of acquiring the actual stock price of each company to be distributed, and an acquired investor's forecast in accordance with the acquisition of new company information. It is characterized by including a means for automatically generating a company information article including a stock price and an actual stock price of the same company as the company related to the investor's expected stock price.

Further, the present invention is an article generation device that automatically generates a company information article based on company information including numerical information corresponding to a plurality of items related to the performance of the company distributed from the information distribution source. A means for detecting that new company information is received by an external information database system that receives the company information to be delivered and stores the forecast company information including the numerical information of the forecast according to the plurality of items. , When the reception of new company information by the external information database system is detected, the new company information and the predicted company information of the same company as the company related to the new company information can be obtained from the external information database system. , A means to compare the numerical information according to multiple items included in the acquired new company information with the numerical information of the forecast according to multiple items included in the acquired forecast company information, and the compared numerical value. Among the information, a means for specifying the item related to the numerical information having the largest value according to the difference between the numerical information, and a means for describing the information corresponding to the specified item in the heading or subject of the company information article. It is characterized by having.

Further, in the present invention, the external information database system accumulates the company information received from the information distribution source as the past company information, and detects the reception of new company information by the external information database system. In the case, in addition to the new company information and the forecast company information, a means for acquiring the past company information of the same company as the company related to the new company information from an external information database system, and the new company information. Numerical information according to a plurality of items included in the above, numerical information of a forecast according to a plurality of items included in the forecast company information, and past numerical information corresponding to a plurality of items included in the acquired past company information. It is characterized by having a comparison means for comparing each item.

Furthermore, in the present invention, the comparison means compares the numerical information included in the new company information with the numerical information which is the past extreme value in the numerical information included in the past company information. It is characterized by being.

In the present invention, the plurality of items include important items, and the numerical information included in the new company information includes both a case where it is larger and a case where it is smaller than the numerical information to be compared. In the comparison of numerical information related to important items, if the numerical information included in the new company information is larger than the numerical information to be compared, it is an evaluation means for performing a positive overall evaluation of the new company information. , The above-mentioned positive overall evaluation is provided with a means for attaching to a company information article.

Further, in the present invention, when the evaluation means includes both the case where the numerical information included in the new company information is larger and the case where the numerical information included in the new company information is smaller than the numerical information to be compared, the comparison relating to the important item is performed. If the numerical information contained in the new company information is smaller than the numerical information to be compared, a negative overall evaluation is performed for the new company information, and the negative overall evaluation is included in the company information article. It is characterized by being provided with means for attaching.
Further, the present invention is characterized in that the important item is ordinary profit.

Furthermore, according to the present invention, the external information database system stores average numerical information obtained by averaging numerical information corresponding to a plurality of items included in corporate information of a plurality of companies, and acquisition of new corporate information. A means of acquiring average numerical information from an external information database system and a means of comparing numerical information corresponding to a plurality of items included in the acquired new company information with the acquired average numerical information for each item. And, based on the comparison, the item relating to the numerical information having the largest value according to the difference between the numerical information and the average numerical information is provided with a means for making the important item.

In the present invention, the external information database system stores average numerical information obtained by averaging numerical information corresponding to a plurality of items included in the corporate information of a plurality of companies, and is used for acquiring new corporate information. In addition, a means of acquiring average numerical information from an external information database system and a means of comparing numerical information corresponding to multiple items included in the acquired new company information with the acquired average numerical information for each item. , Based on the comparison, the means for specifying the order of the items related to the numerical information of the comparison target according to the order of the values according to the difference between the numerical information and the average numerical information, and the items related to each item according to the order of the specified items. It is characterized by providing a means for describing information and generating a company information article.

Further, the present invention relates to an article generation method in which an article generator automatically generates a company information article based on company information including numerical information corresponding to a plurality of items related to the performance of the company distributed from the information distribution source. The article generator receives new company information distributed from an information distribution source and stores new company information by an external information database system that stores forecast company information including numerical information of prediction according to the plurality of items. When the step of detecting that the information is received and the reception of new company information by the external information database system are detected, the new company information and the predicted company information of the same company as the company related to the new company information are displayed. Steps to be acquired from an external information database system and numerical information according to multiple items included in the acquired new company information, numerical information and items of forecast according to multiple items included in the acquired forecast company information Company information article about the step to compare for each, the step to specify the item related to the numerical information with the largest value according to the difference between the numerical information among the compared numerical information, and the information according to the specified item. It is characterized by having a step described in the heading or subject.

Further, the present invention is for causing a computer to perform a process of automatically generating a company information article based on company information including numerical information corresponding to a plurality of items related to the performance of the company distributed from the information distribution source. In the computer program, a new information database system is used to receive the company information distributed from the information distribution source to the computer and store the forecast company information including the numerical information of the forecast according to the plurality of items. When the step of detecting the reception of company information and the reception of new company information by an external information database system are detected, the new company information and the forecast company information of the same company as the company related to the new company information , And the numerical information according to multiple items included in the acquired new company information, and the numerical information of the forecast according to multiple items included in the acquired forecast company information. And the step of comparing each item, the step of identifying the item related to the numerical information with the largest value according to the difference between the numerical information among the compared numerical information, and the information according to the specified item, the company It is characterized by performing the steps described in the heading or subject of an information article.

In the present invention, the numerical information according to a plurality of items included in the new company information is compared with the numerical information of the forecast according to the plurality of items included in the forecast company information (performance forecast information) for each item. Then, the item related to the numerical information with the largest value according to the difference is specified, so among the multiple items included in the distributed company information, the item with the largest difference compared to the forecast is determined. Will be. Then, by describing the item in the headline or subject of the article, it becomes possible to provide an article in which the item that is characteristic in comparison with the prediction is conspicuous.

In the present invention, numerical information corresponding to a plurality of items included in new company information is compared for each item with each numerical information corresponding to a plurality of items included in past company information in addition to prediction. Since the item related to the numerical information according to the value with the largest difference is specified, the item with the largest difference compared to the forecast or the past is divided among the multiple items included in the distributed company information. Since it is issued and the item is described in the headline or subject of the article, it becomes possible to provide an article in which the item that is characteristic in the prediction or the comparison with the past is conspicuous.

In the present invention, numerical information corresponding to a plurality of items included in the new corporate information is averaged numerical information corresponding to a plurality of items included in the corporate information of the plurality of companies, and the average numerical information and each item are averaged. Compared to, the order of the items with the largest difference between the two is specified, and the information related to each item is described according to the order to generate the article, so the article content in which the items with the largest difference from other companies are arranged in order. Therefore, it will be possible to provide articles with a structure that makes it easy to understand the differences in the performance of other companies that investors are paying close attention to.

In the present invention, since an article including the theoretically calculated theoretical stock price of the company and the actual stock price of the company is generated, the difference between the actual stock price and the theoretical stock price, which can be used as a reference for future investment, etc. It will be possible to provide articles that are easy for investors to understand.

In the present invention, since an article including the investor's expected stock price of a company calculated by aggregating the stock prices predicted by each of a plurality of investors and the actual stock price of the company is generated, a large number of investors It will be possible to provide articles that investors can easily understand the difference between the expected stock price that represents collective knowledge and the actual stock price, and investors can use such articles as materials for studying future investment. ..

In the present invention, since the numerical information included in the new corporate information is compared with the numerical information that becomes the past extreme value (maximum value or minimum value), the business performance of the company has reached a record high or a record low. It will be possible to express such things in articles, and it will be possible to simply express the current state of corporate performance in the flow of business performance trends of a series of companies. As for the past range, it is preferable to keep the 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 situation. Is.

In the comparison of each item, when the numerical information included in the new company information includes both a large case (plus case) and a small case (minus case) with respect to the comparison target, in general, It is difficult to evaluate whether or not the content of the newly announced corporate information is good, but in the present invention, important items are defined among a plurality of items, and the comparison results are obtained as described above. When it is difficult to evaluate due to a mixture of positive and negative cases, compare only the important items. And if the comparison result of the important items is positive, the overall evaluation of the newly announced corporate information will be positive, so even when the content of the corporate information is mixed with good and bad for each item and it is difficult to evaluate. By evaluating from the viewpoint of important items, it is possible to show investors, etc. certain guidelines in an easy-to-understand manner.

Further, in the present invention, when positives and negatives are mixed in the comparison result as described above, if the comparison result is negative in the comparison of important items, the overall evaluation of new corporate information is negative. Therefore, even when it is difficult to evaluate the content of corporate information, it is possible to provide investors with the content that makes it possible to understand the overall evaluation at a glance by evaluating from the viewpoint of important items. Become.

Further, in the present invention, the ordinary profit is set as an important item, and the ordinary profit generally represents the business performance of the company most prominently. Therefore, by setting the important item as the ordinary profit, the ordinary profit is the axis. As a result, it will be easier to properly evaluate the newly announced corporate information as a whole.

In the present invention, an average numerical value obtained by averaging numerical information corresponding to a plurality of items included in newly announced corporate information and numerical information corresponding to a plurality of items included in the corporate information of a plurality of companies. Since the item with the largest difference between the information and each item is set as the important item, it will be possible to evaluate the new company information as a whole by focusing on the item with the largest difference from other companies. ..

In the present invention, among a plurality of items included in the new corporate information to be distributed, the item that is the largest, which is different from the prediction, is described in the headline or the subject of the article. It is possible to provide investors, etc. with an automatically generated article that highlights the items that are characteristic in comparison with.

In addition to forecasting, each numerical information corresponding to multiple items included in the past company information is also compared for each item, and the item related to the numerical information according to the value with the largest difference is identified, so the past to the future. By comparing over a wide range of time, the item with the largest difference is identified and described in the headline or subject of the article, so that in corporate performance, both past and future. It is possible to provide an article in which characteristic items stand out when viewed.

Furthermore, in the present invention, when the record high or record low is updated among a plurality of items included in the new corporate information to be distributed, that fact is described in the headline or subject of the article. , It is easy to tell investors that the current corporate performance is at its peak.
Furthermore, in the present invention, the content of the article is arranged in the order of items that are significantly different from other companies, so that it is easy to understand the content that stands out when compared with other companies that investors and others are paying close attention to. Can provide articles.

In the present invention, an article including the theoretically calculated theoretical stock price of a company and the actual stock price of the company is generated. Therefore, as an investment judgment material, an article in which the actual stock price and the theoretical stock price can be compared is used. Can be provided.
Further, in the present invention, since an article including the investor's expected stock price calculated by aggregating the stock prices predicted by each of a plurality of investors and the actual stock price is generated, it is based on the collective knowledge of a large number of investors. You can provide an article that allows you to compare the expected stock price with the actual stock price.

In the present invention, if an important item is defined among a plurality of items and the comparison result for each item is a mixture of positive and negative results, the newly announced company information is based on the comparison result in the important item. Since the article shows that the overall evaluation of is positive or negative, it is possible to present an evaluation from the viewpoint of important items as a certain guideline for investors.
Further, in the present invention, by setting the important item to the ordinary profit in which the business performance of the company is remarkably shown, the newly announced company information can be appropriately evaluated.
Further, in the present invention, it is possible to make an overall evaluation of newly announced corporate information by focusing on the items that are most different from other companies.

It is the schematic which shows the overall structure of the article providing system including the article generation system which concerns on embodiment of this invention. It is a schematic diagram which shows the internal structure of an information DB system. It is a block diagram which shows the main internal structure of an article generator. It is a chart which shows an example of the contents of the quarterly settlement table part included in the evaluation standard table. It is a chart which shows an example of the contents of the full-year settlement table part included in the evaluation standard table. It is a flowchart which shows a series of processing procedures concerning an article generation method. It is a schematic diagram which shows the example of the company information article which was automatically generated for the announced financial statements of company A. It is a schematic diagram which shows the example of the company information article which was automatically generated for the announced financial statements of company B. (A) is a schematic diagram showing an example of a company information article automatically generated for the announced revision of the earnings forecast of company C, and (b) is a schematic diagram for the announced revision of the earnings forecast of company D. It is the schematic which shows the example of the automatically generated company information article. It is a schematic diagram which shows the example of the company information article automatically generated for the announced dividend adjustment of company E. (A) is a schematic diagram showing an example of stock price information including the theoretical stock price of company A, (b) is a schematic diagram showing an example of stock price information including the investor forecast stock price of company A, and (c) is announced. It is a schematic diagram which shows an example of the comprehensive evaluation (overall evaluation) with respect to the settlement of accounts of company A. It is a schematic diagram which shows the example of the company information article visualized (designed). (A) and (b) are schematic views showing an example of the explanatory text generated by the prior art. (A) is a schematic diagram showing an example of an article automatically generated by the prior art for the announced financial statements of company A, and (b) is automatically generated by the prior art for the announced financial statements of company B. It is the schematic which shows the example of the article.

FIG. 1 is a schematic view showing an overall configuration of an article providing system 1 including an article generating system 5 according to the present invention. The article providing system 1 automatically generates a text article that is more advanced than the conventional one by the article generation system 5, and makes it possible to provide each user's user terminals T1, T2, T3, T4, etc. through a network. The article providing system 1 is a stock exchange system 2 and each of which distributes various company 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). It has a stock price distribution system 3 that distributes fluctuating stock prices of listed companies at any time.

Further, 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 provided with a plurality of types of DBs (databases) that store independently constructed data archives. In this embodiment, the corresponding settlement DB15, past settlement DB16, performance forecast DB17, and theoretical stock price DB18 , And the individual investor stock price forecast DB19. The information DB system 10 acquires various types of information (corresponding to the company information announced by listed companies) distributed by the stock exchange system 2 at any time, and the article generator 20 monitors the acquisition of such company information. (Detection), the article generation device 20 automatically generates an article based on the above-mentioned announced company information, triggered by the information DB system 10 acquiring the company information. Hereinafter, the contents of the present invention will be described in detail with a focus on the information DB system 10 and the article generation device 20 constituting the article generation system 5.

FIG. 2 outlines the internal configuration of the information DB system 10, and the DB management device 11 is connected to each of the above-mentioned DBs 15 to 19 by an internal connection line 10a. The DB management device 11 manages and controls each of the DBs 15 to 19, receives (acquires) information and data distributed (provided) from an external system or the like, and incorporates the received information or the like into the contents thereof. It sorts the data according to the situation and stores it in each of the required DBs 15 to 19, and also reads the information stored in each DB 15 to 19 in response to an external request and sends the read information to the requester. Perform processing, etc.

The DB management device 11 has a control unit 11a and a communication unit 11b inside, and the control unit 11a is substantially responsible for each process of the DB management device 11 described above, and controls each of these processes. Do. The control unit 11a has a memory (storage unit) inside, so that information, data, and the like associated with the processing can be temporarily stored. Further, the communication unit 11b communicates with the external network and each DB 15 to 19 under the control of the control unit 11a, and each information, data, etc. for each DB 15 to 19 by the communication of the communication unit 11b. Can be sent and received.

The corresponding settlement DB 15 possessed by the information DB system 10 is a database that stores company information announced by each listed company distributed by the above-mentioned securities exchange system 2 in association with each company, and the DB management device 11 is a securities. When the company information distributed from the exchange system 2 (corresponding to the information distribution source) is received and acquired, the acquired company information is stored (stored) in the corresponding settlement database 15. The company information distributed from the stock exchange system 2 is in the predetermined standardized file format (XBRL format) described above, and the numerical information included in the XBRL format file is in units of 1 million yen. (For example, 10 billion yen is written as "10000"). Therefore, as it is, it is difficult to analyze the company information and generate sentences at a later stage. Therefore, the DB management device 11 stores the numerical information included in the acquired XBRL format company information in the corresponding settlement DB 15. , Performs numerical unit conversion processing, and saves (stores) the converted contents in the corresponding settlement database 15 (for example, the above-mentioned numerical information of "10000" is converted into the numerical unit of "10000000000000". ).

Further, the DB management device 11 makes it possible to detect from the outside that the company information delivered from the stock exchange system 2 has been received, and when the company information is received, the company information of any company is displayed at any time. An information signal indicating whether or not it has been received is issued, and the issued information signal is transmitted to the outside. The company information distributed by the Stock Exchange System 2 includes financial statements (full-year, first-half, second-half, quarterly cumulative, quarterly, etc.) announced by each listed company, earnings forecasts, revisions to earnings forecasts, dividend forecasts, and dividends. There are revisions to forecasts, etc., and corporate information such as financial results, earnings forecasts, revisions to earnings forecasts, dividend forecasts, revisions to dividend forecasts, etc. is numerical information (%) according to multiple items related to corporate performance. Information indicating a numerical value corresponding to a ratio or a monetary unit or a non-unit numerical value), and is distributed to the information DB system 10 according to the timely disclosure information browsing service operated by the stock exchange system 2.

If the company information is a settlement of accounts, there are multiple items such as ordinary income, operating income, sales, and net income, and the numerical information of each company for each of these items (numerical information indicating the numerical value of ordinary income). , Numerical information indicating the numerical value of sales, numerical information indicating the numerical value of operating profit, numerical information indicating the net profit, etc.) will be included in the corporate information. In addition, the distribution (announcement) of corporate information from the stock exchange system 2 is generally announced after the closing (15:00) or around 15:30, and the article generation system 5 of the present invention is described in this way. It is possible to provide generated articles at almost the same time as the announcement time.

If the company information is revised (revision of earnings forecast, revision of dividend forecast, etc.), items related to profit such as revision of ordinary income and revision of final profit will be included. , Numerical values (numerical information) indicating dividend adjustments are included.

The past settlement DB 16 possessed by the information DB system 10 is a database that stores past company information (past company information), and the past is stored by archiving the company information stored in the corresponding settlement DB 15 described above. It is corporate information. The storage process in the past settlement DB 16 is linked to the storage process of the corresponding settlement DB 15, and when the DB management device 11 stores the company information in the corresponding settlement DB 15, the stored company information is stored in the past settlement DB 16. Is also stored. Therefore, only the most recently announced new company information is stored in the relevant settlement DB15, and the old company information is deleted at any time (for example, about a few weeks after the announcement). Elapsed information, etc.). On the other hand, in the past settlement DB16, the company information once stored is accumulated as the past company information for each company, and is a suitable database for confirming the past performance and the like. As for the company information stored in the past settlement DB 16, the numerical information included in the above-mentioned XBRL format has been converted.

Further, the DB management device 11 performs a process of calculating an average value corresponding to each of a plurality of items with respect to the numerical information included in the past company information for each company accumulated in the past settlement DB 16, and each calculated value. The average value for each item is also stored in the past settlement DB 16 as average numerical information. This average value calculation process is performed with the content of calculating the average value of other companies excluding one company. For example, assuming that there are companies A to Z as companies that acquire company information, the calculation process of the average value excluding company A is performed using the company information of companies B to Z, excluding company B. The calculation process of the average value is performed using the company information of companies A, C to Z, and the calculation process of the average value excluding company C is performed using the company information of companies A, B, D to Z. Further, the calculation process of the average value excluding the company Z is performed using the company information of the companies A to Y.

Then, the DB management device 11 stores the average value (calculated average value) excluding company A in the past settlement DB 16 in association with company A, and similarly, the average value excluding company B (calculated). The average value) is stored in the past settlement DB16 in association with company B, and the average value (calculated average value) excluding company C is stored in the past settlement DB16 in association with company C, excluding company Z. The average value (calculated average value) is stored in the past settlement DB 16 in association with Company Z. By calculating the average value in this way and storing it in association with each company, the stored average value can be used when comparing the performance of a certain company with the average value of another company. In addition, the calculation process of such an average value is performed in a predetermined period unit (for example, a monthly unit such as every one month or three months, a year unit such as every six months, one year, three years, or every five years). The DB management device 11 performs according to the announcement of business results from the company, and the average value calculated and stored is a 3-month average value, a 6-month average value, a 1-year average value, a 3-year average value, or a 5-year average value. A plurality of average values may be prepared as the average value.

Further, the performance forecast DB 17 possessed by the information DB system 10 is a database that stores performance forecast information (corresponding to forecast company information) including numerical information that forecasts the performance of each listed company in association with each company. The performance forecast information to be stored is generated by the operating entity or the forecasting company of the information DB system 10, and like the corporate information distributed from the stock exchange system 2, a plurality of business performance related to the company. Includes numerical information of forecasts according to the items in. When the earnings forecast information generated from the device of the forecasting company is sent to the information DB system 10, the DB management device 11 receives and acquires the earnings forecast information, and the earnings forecast DB 17 is divided for each company. The process of storing separately is performed. In addition, such earnings forecast information also conforms to the above-mentioned predetermined standardized file format, and the numerical unit of the numerical information is also a general unit (the first digit starts from the 1st digit). unit).

Further, the theoretical stock price DB 18 possessed by the information DB system 10 associates the stock price of each listed company with the theoretical stock price (theoretical stock price information) indicating the numerical value of the stock price calculated by using a mathematical formula based on a predetermined theory or the like. It is a stored database. As an example, such theoretical stock price information is calculated and generated at any time by an entity that manages and operates the article generation device 20, an entity that manages and operates the information DB system 10, or an external entity. When the theoretical stock price information is sent to the information DB system 10 from the device (the device having the calculated theoretical stock price information) by the business entity, the DB management device 11 receives and acquires the theoretical stock price information, and the theory The stock price DB 18 is processed to be stored separately for each company.

Finally, the individual investor stock price forecast DB 19 included in the information DB system 10 is a database that stores individual investor forecast stock price information indicating the forecast stock price of each company by a plurality of individual investors. The individual investor forecast stock price information to be stored is provided by an entity that aggregates the forecast stock price of each individual investor, and the device possessed by this entity aggregates the forecast stock price of each individual investor. By performing statistical processing (averaging processing, etc.), stock price information based on individual investor forecasts (individual investor forecast stock price information including numerical information that forecasts the stock price of each company) is generated and generated. The individual investor forecast stock price information is provided (transmitted) to the information DB system 10 and the like. An example of an entity that aggregates such expected stock prices is an entity that manages and operates the article generation device 20, an entity that manages and operates the information DB system 10, an external entity, and the like. Further, when the information DB system 10 receives the individual investor forecast stock price information sent from the device of the business entity that aggregates the forecast stock price, the DB management device 11 receives the received individual investor forecast stock price information for the individual investor. The stock price forecast DB 19 is processed to be associated and stored for each company.

The DB management device 11 of the information DB system 10 performs management control of each of the DBs 15 to 19 described above, and when various information about each company is received from the outside of the information DB system 10, the contents of the information are detected. A process is performed in which the received information is stored (stored) in any of the DBs 15 to 19 according to the contents so that the received information can be identified for each company. Further, when the DB management device 11 of the information DB system 10 receives a request for information (company information request) about a specific company from the article generation device 20 that automatically generates articles, each DB 15 is associated with the company. Various information stored in ~ 19 (various information about the company to be requested) is read from each DB 15 ~ 19 and sent (provided) to the requesting article generation device 20 (various information is). Each is sent as a file).

Specifically, when the DB management device 11 receives an information request about the listed company A, the DB management device 11 distributes the latest company information of the company A (stock exchange system 2) from the corresponding settlement DB15. Read the latest settlement of accounts, earnings forecast, revision of earnings forecast, dividend forecast, revision of dividend forecast, etc.), and from the past settlement DB16, the past company information of company A (past company information file) and Information showing the average value of other companies excluding company A (average numerical information) is read out, company A's performance forecast information (performance forecast information file) is read from the performance forecast DB17, and the theoretical stock price of company A is read from the theoretical stock price DB18. Information (theoretical stock price information file) is read out, and the individual investor forecast stock price information (individual investor forecast stock price information file) of company A is read out from the individual investor stock price forecast DB19, and various information (each information) of company A read out. File) will be sent to the article generator 20. In this embodiment, the company information for the past read from the past settlement DB 16 is the information for the past 5 years (information on the settlement, business performance, dividends, etc. announced within the past 5 years), but the system Depending on the specifications and the like, appropriate past information may be applied within the range of the past 10 years. Further, the information DB system 10 corresponds to an information database system external to the article generation device 20.

FIG. 3 is a block diagram showing a main internal configuration of the article generation device 20, which is the central device of the present invention. The article generation device 20 of the present embodiment is constructed by a general server computer (server device), but a system may be constructed by combining a plurality of server devices or the like by performing distributed processing or the like. Of course it is possible. The article generation device 20 automatically generates articles based on various information and data archived in the above-mentioned information DB system 10 by an original program (AI: artificial intelligence) triggered by the announcement of company information.

The article generator 20 is a device in which various devices and the like are connected to the MPU 20a that performs overall control and various processes by an internal connection line 20h, and the communication module 20b, RAM 20c, ROM 20d, and inputs are connected to the various devices and the like. There are an interface 20e, an output interface 20f, a storage unit 20g, and the like.

The communication module 20b is a communication device corresponding to a connection module with a network, and conforms to a required communication standard (for example, a LAN module). The communication module 20b is connected to a network (including internal, external, or other communication networks, etc.) via a required communication device (not shown, for example, a router or the like), and includes a stock price distribution system 3, information. It enables communication with the DB system 10 (corresponding to an external information database system for the article generation device 20) and user terminals T1, T2, T3, and the like. Since the stock price of each company is distributed from the stock price distribution system 3 during the time when the stock market is open, the distributed stock price is received by the communication module 20b and received by the RAM 20c, the storage unit 20g, or the like. The distribution stock prices are stored in sequence.

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, a mouse, and the like that receive operation instructions and the like from the administrator of the article generation device 20. A display 22 (display output device) is connected to the output interface 20f, and the contents associated with the processing of the MPU 20a are output to the display 22, and the administrator of the article generation device 20 can check the current processing contents and the like. I am trying to do it.

The storage unit 20g stores a program, a table, etc., stores an OS program P1 and an article generation program P2 as a program, and stores a created article table 24, an evaluation standard table 25, etc. as a table. .. In order to install the article generation program P2 in 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 in the storage unit 20g through the storage medium.

The OS program P1 in the program stored in the storage unit 20g defines various processes according to the operating system for the server computer, and the MPU 20a performs the processes based on the specified contents to generate an article. The device 20 fulfills each function as a server computer (server device).

Further, the article generation program P2 stored in the storage unit 20g defines various processes for the MPU 20a to automatically generate articles as artificial intelligence, and the MPU 20a defines the information DB system 10 in accordance with the provisions of the article generation program P2. Receiving and monitoring corporate information by, analysis and evaluation of corporate information acquired from information DB system 10, identification of article headings based on the results of analysis and evaluation, identification of important items, identification of explanation order of article contents, etc. Act as a means of doing. In addition, in order to create sentences related to articles, the article generation program P2 also includes the processing contents according to the sentence creation engine, and the MPU 20a is made into AI (artificial intelligence) for sentence creation by this sentence creation engine. Also works as. The details of such an article generation program P2 will be described based on the flowchart shown in FIG. 6 described later, and the created article table 24 and the evaluation standard table 25 will be described before that.

The created article table 24 stores articles (company information articles) automatically created by the MPU 20a according to the provisions of the article generation program P2 for each company corresponding to the articles. In addition, the evaluation standard table 25 stores conditions and the like related to various comparisons used for analysis and evaluation processing of corporate information.

FIG. 4 shows an example of the contents of the quarterly settlement table unit 25a included in the evaluation standard table 25. The quarterly settlement table section 25a shown in FIG. 4 is used when the announced new company information is used for quarterly settlement, and points indicating the evaluation given as the analysis result (value of evaluation common to different items). Specify the numerical value of). The quarterly settlement table section 25a provides examples of analysis and evaluation items such as "(1) Comparison of revised earnings forecast and conventional earnings forecast when quarterly settlement is revised ant" and "(2) When quarterly settlement is revised ant". "Comparison of revised dividend forecast and conventional dividend forecast", "(3) Comparison of actual figures and disclosed forecasts when quarterly financial results are not revised", "(4) Comparison of financial results and past results", "(5) ) "Comparison of financial results and record highs" and "(6) Comparison of profit margins and past results" etc. are included.

"(1) Comparison of revised earnings forecast and conventional earnings forecast when quarterly financial results are revised" is applied when the announced new company information is revised quarterly financial results (correction of earnings forecast, etc.) It is a matter, and the value of points given (or subtracted) according to the result of comparison between the revised earnings forecast and the conventional (before revision) earnings forecast is specified. Specifically, if the revision rate of the earnings forecast is 15% or more and less than 30% (15% ≤ revision rate <30%), 1 point is given (plus 1 point), and the earnings forecast is based on the conventional one. If the revision rate is 30% or more (30% ≤ correction rate), 2 points are given (plus 2 points), while if the revision rate of the earnings forecast is -15% or less and less than 30% (plus 2 points). -30% <correction rate ≤ -15%), subtract 1 point (-1 point), and if the revision rate of the earnings forecast is -30% or less (-30% ≥ correction rate), 2 points Subtraction (minus 2 points) is stipulated.

"(2) Comparison of revised dividend forecast and conventional dividend forecast when quarterly financial results are revised" is also a matter applied when the announced company information is revised quarterly financial results (correction of dividend forecast, etc.). Yes, it stipulates the value of points to be given (or subtracted) according to the result of comparison between the revised dividend forecast and the conventional (before revision) dividend forecast. Specifically, if the revision rate of the dividend forecast is 15% or more and less than 30% (15% ≤ revision rate <30%), 1 point is given (plus 1 point), and the dividend forecast is compared to the conventional one. If the correction rate is 30% or more (30% ≤ correction rate), 2 points are given (plus 2 points). On the other hand, if the revision rate of the dividend forecast is -15% or less and less than 30% (-30% <correction rate ≤ -15%), 1 point is subtracted (-1 point). If the correction rate is -30% or less (-30% ≥ correction rate), it is stipulated that 2 points should be subtracted (minus 2 points).

"(3) Comparison of actual figures and disclosed forecasts when quarterly financial results are not revised" is a matter that is applied when the announced company information is quarterly financial results without revision, and the announced business results. It stipulates the value of points given (or subtracted) according to the result of the progress rate obtained by comparing the notification with the disclosed earnings forecast. Specifically, if the progress rate is 80% or more with respect to the disclosed forecast (80% ≤ progress rate), 1 point is given (plus 1 point), while the progress rate is less than 60%. If so, it is stipulated that 1 point should be subtracted (minus 1 point).

"(4) Comparison between financial results and past results" is a matter that is applied when the announced company information is quarterly financial results, and the announced financial results and past results (results of the latest period) ) Is given (or subtracted) according to the comparison result. Specifically, if the announced financial results are more than 15% of the past results and less than 30% (past results + 15% ≤ financial results <past results + 30%), 1 point will be awarded. If the announced financial results are 30% or more of the past results (plus 30% of the past results ≤ financial results), 2 points will be given (plus 2 points), while the announcement will be made. If the settlement value is less than -15% of the past record and less than 30% (-30% of the past record <Financial statement ≤ -15% of the past record), subtract 1 point (-1 point). If the announced financial results are -30% or less of the past results (-30% of the past results ≥ financial results), it is stipulated that 2 points should be subtracted (minus 2 points).

"(5) Comparison between financial results and record highs" is also a matter that is applied when the announced company information is quarterly financial results or financial forecasts, and the announced financial results are regarded as the past high results. In contrast, it is stipulated that points will be awarded depending on whether the record high performance is updated (exceeded) or whether the expected financial results included in the announcement are expected to update the past high performance. It has become a thing. Specifically, if the announced financial results are the highest ever, give 1 point (plus 1 point), and expect the announced forecast to be the highest ever. If so, it is stipulated that 1 point will be given (plus 1 point).

"(6) Comparison of profit margin and past performance" is a matter that is applied when the announced company information is in the case of quarterly settlement of accounts, and the profit margin included in the announced company information and the past performance (for example, The value of points given (or subtracted) according to the result of comparison with the profit margin of the latest period or the profit margin of the same period of the previous year, etc.) is specified. Specifically, if the profit margin included in the announced company information is + 15% or more and less than 30% of the past performance (+ 15% of the past performance ≤ profit ratio <+ 30% of the past performance) 1 point is given (plus 1 point), and if the announced profit margin is 30% or more of the past performance (past performance plus 30% ≤ profit margin), 2 points are given (plus 2 points). ) On the other hand, if the profit margin included in the announced company information is -15% or less and less than 30% of the past performance (-30% of the past performance <profit margin ≤ -15% of the past performance). 1 point is subtracted (minus 1 point), and if the announced profit margin is -30% or less of the past performance (-30% of the past performance ≥ profit margin), 2 points are subtracted (minus 2 points). ) Is stipulated.

FIG. 5 shows an example of the contents of the full-year settlement table unit 25b included in the evaluation standard table 25. The full-year settlement table section 25b shown in FIG. 5 defines the numerical values of points (evaluation values) given as a result of analysis and evaluation, etc., because the company information to be analyzed and evaluated is used in the case of full-year settlement. To do. The full-year financial statements table section 25b provides analysis and evaluation items such as "(10) Comparison of current term earnings forecast and past results", "(11) Comparison of settlement figures and record highs", and "(12) Previous term actual results". Includes "Comparison of forecasts", "(13) Comparison of financial results in units of 3 months with past results", and "(14) Comparison of profit margins with past results".

In "(10) Comparison of current term earnings forecast and past results", the current term earnings forecast (for example, ordinary income forecast) included in the announced corporate information (financial statements) is used as the past performance value (for example, the current term for the most recent period). The value of points to be given is specified according to the result of comparison with (profit). Specifically, if the earnings forecast for the current term included in the announced corporate information is + 30% or more and less than 50% of the past actual value (plus 30% of the past actual value ≤ current term earnings forecast <past actual value) 1 point (plus 1 point), and if the current term earnings forecast is 50% or more of the past earnings forecast (plus 50% of the past earnings forecast ≤ this term earnings forecast), 2 points Granted (plus 2 points), on the other hand, if the current term earnings forecast is -30% or less of the past actual value and less than 50% (-50% of the past actual value <current term earnings forecast ≤ -30 of the past actual value) %), 1 point is subtracted (minus 1 point), and if the current term earnings forecast is -50% or less of the past actual value (-50% of the past actual value ≥ current term earnings forecast), 2 points are subtracted. (Minus 2 points) is stipulated.

"(11) Comparison between financial results and record highs" is the same as in the case of "(5) Comparison between financial results and record highs" described above, and is included in the announced corporate information (settlement). Compare the numerical value with the past highest performance (past highest settlement value), and whether the past highest performance is updated (exceeded) or the announced settlement forecast value included in the announced company information (financial forecast) It stipulates that points will be awarded depending on whether or not the highest performance in the past is expected to be updated. Specifically, if the financial results included in the announced company information update the highest record in the past, 1 point will be given (plus 1 point), and the announced forecast of financial results will be the highest record in the past. If renewal is expected, it is stipulated that 1 point will be given (plus 1 point).

In "(12) Comparison of previous term actual value and forecast", the previous term actual value (for example, the previous term's ordinary income) included in the announced corporate information (settlement) is used as the next term's forecast value (for example, next term's ordinary income forecast). The value of points to be given is specified according to the result of comparison with (value). Specifically, if the previous term actual value included in the announced company information is + 30% or more and less than 50% of the next term forecast value (plus 30% of the next term forecast value ≤ previous term actual value <next term forecast value) (Plus 50%), 1 point is given (plus 1 point), and if the previous term actual value is 50% or more of the next term forecast value (next term forecast value plus 50% ≤ previous term actual value), 2 points Granted (plus 2 points), on the other hand, if the previous term actual value is -30% or less and less than 50% of the next term forecast value (-50% of the next term forecast value <previous term actual value ≤ next term forecast value -30) %), 1 point is subtracted (minus 1 point), and if the previous term actual value is -50% or less of the next term forecast value (-50% of the next term forecast value ≥ previous term actual value), 2 points are subtracted. (Minus 2 points) is stipulated.

"(13) Comparison between 3-month financial results and past results" is the previous year's 3-month financial results (for example, 3-month financial results) included in the announced company information (settlement). The value of points to be given is specified according to the result of comparison with the actual results of the same period (for example, settlement figures). Specifically, if the three-month financial results included in the announced company information are + 30% or more and less than 50% of the actual value for the same period of the previous year (plus the actual value for the quarter of the same period of the previous year). 30% ≤ 3 months settlement value <+ 50% of the actual value for the same period of the previous year), 1 point is given (plus 1 point), and the 3 month unit settlement value is 50 compared to the actual value for the same period of the previous year. If it is% or more (plus 50% of the actual value for the same period of the previous year ≤ the settlement value in units of 3 months), 2 points will be given (plus 2 points), while the settlement value in units of 3 months will be the actual value for the same period of the previous year. On the other hand, if it is -30% or less and less than 50% (-50% of the actual value for the same period of the previous year <Financial value for 3 months ≤ -30% of the actual value for the same period of the previous year), 1 point is subtracted (minus 1). (Points) If the settlement value for 3 months is -50% or less of the actual value for the same period of the previous year (-50% of the actual value for the same period of the previous year ≥ the settlement value for 3 months), 2 points are subtracted ( Minus 2 points) is stipulated.

"(14) Comparison of profit margin with past portion" is the same as in the case of "(6) Comparison of profit margin with past performance", the profit margin included in the announced corporate information and the past portion. The value of points given (or subtracted) according to the result of comparison with the actual result (profit margin in the previous term) is specified. Specifically, if the profit margin included in the announced company information is + 15% or more and less than 30% of the past performance (+ 15% of the past performance ≤ profit ratio <+ 30% of the past performance) 1 point is given (plus 1 point), and if the announced profit margin is 30% or more of the past performance (past performance plus 30% ≤ profit margin), 2 points are given (plus 2 points). ) On the other hand, if the profit margin included in the announced company information is -15% or less and less than 30% of the past performance (-30% of the past performance <profit margin ≤ -15% of the past performance). 1 point is subtracted (minus 1 point), and if the announced profit margin is -30% or less of the past performance (-30% of the past performance ≥ profit margin), 2 points are subtracted (minus 2 points). ) Is stipulated.

The conditions (including numerical values associated with the conditions) related to the points shown in the quarterly settlement table section 25a in FIG. 4 and the full-year settlement table section 25b in FIG. 5 are merely examples, and other conditions may be applied. Of course it is possible. For example, the conditions for dividends are explained in the case of the quarterly settlement of accounts as "(2) Comparison of the revised dividend forecast and the conventional dividend forecast when the quarterly settlement of accounts is the revised ant" in the quarterly settlement table section 25a of FIG. However, even in the quarterly and full-year financial results without revision, if the dividend announced this time is an increase of 5 yen or more and less than 10 yen, it will be an increase of 1 point or 10 yen or more. If so, the contents such as plus 2 points, minus 1 point if the down is 5 yen or more and less than 10 yen, and minus 2 points if the down is 10 yen or more are also specified.

Further, the table portions 25a and 25b shown in FIGS. 4 and 5 are a part included in the evaluation standard table 25, and include comparison conditions with other objects and the like. For example, as other comparison targets, in addition to past results (including record high results or record low results, etc.), forecast values for the next fiscal year, etc., the case of using the average value (average numerical information) of multiple companies for each item is also evaluated. Even when it is included in the standard table 25 and this average numerical information is used as a condition, as in the case of the above-mentioned past results or forecast values for the next fiscal year, it is possible to compare with the numerical information included in the announced company information. Includes provisions for granting plus 1 or 2 points, or minus 1 or 2 points. Further, the evaluation standard table 25 has a table section including contents that specify the conditions related to the points as shown in FIGS. 4 and 5 even when the announced company information is a dividend forecast or a revision of the dividend forecast. And. Next, the article generation program P2 stored in the storage unit 20g will be described in detail.

FIG. 6 is a flowchart explaining each control process performed by the MPU 20a of the article generator 20 defined by the article generation program P2, and the content of the processing procedure shown in this flowchart corresponds to the article generation method. Hereinafter, the processing contents of the MPU 20a defined by the article generation program P2 will be described with reference to the flowchart of FIG.

First, the article generation device 20 (MPU 20a) monitors the status of the information DB system 10 according to the provisions of the article generation program P2, and is distributed from the stock exchange system 2 by the information DB system 10 (DB management device 11). Detects whether company information has been received (S1). In this detection, when the information DB system 10 receives the company information distributed from the stock exchange system 2 as described above, the DB management device 11 sends an information signal indicating when the company information of which company is received. The article generation device 20 (MPU 20a) performs the above detection by transmitting and monitoring the presence / absence of reception of company information by the information DB system 10 based on the presence / absence of this information signal.

When the information DB system 10 has not detected that the company information has been received (S1: NO), the article generator 20 (MPU20a) is in a state of waiting for detection, while the information DB system 10 has received the company information. (S1: YES), the article generation device 20 (MPU 20a) requests the information DB system 10 (DB management device 11) for various information of the company related to the company information for which reception is detected (company information request). (S2).

Then, the article generation device 20 (MPU 20a) receives the company information (files of each company information) sent from the information DB system 10 (DB management device 11) in response to the transmission of the above-mentioned company information request. Acquire (S3). For example, when the company information request about company A is transmitted at the stage of S2 above, the company information of company A (new company information announced by company A whose reception is detected) and the company information of company A in the past are sent. , Average numerical information of multiple other companies (excluding company A), business forecast information of company A (forecast company information), theoretical stock price information of company A, individual investor forecast stock price information of company A, etc. as corporate information The article generation device 20 receives and acquires the information from the information DB system 10. The article generation device 20 (MPU 20a) temporarily stores the acquired company information in the RAM 20d or the storage unit 20g in order to perform subsequent processing.

When the article generator 20 (MPU20a) acquires the requested company information, it performs processing such as calculation and comparison on the acquired company information (numerical information included in the company information) (S4). As a specific example of the contents of the calculation, if the acquired company information is based on the announcement of quarterly financial results, the progress rate of the announced actual figures can be calculated with respect to the earnings forecast, or at the same time in the past. If the announced company information is subject to revision, the revision rate is calculated, and the evaluation standard table 25 (each table part) shown in FIGS. 4 and 5 described above is performed. The points specified in 25a and 25b) are also calculated.

In addition, as the contents of comparison, etc., the comparison between the announced performance value (settlement, profit margin, etc.) and the past performance value (past settlement, profit margin, etc.), and each item of newly announced company information Comparison of the corresponding performance value (settlement and profit margin, etc.) with the performance forecast value (forecast settlement and profit margin) according to each item included in the performance forecast information, and the announced performance forecast value and the past actual value Comparison, comparison of numerical information and average numerical information according to each item of newly announced company information for each item, comparison of progress rate obtained by the above calculation with actual progress rate and expected progress rate, There is a comparison between the actual progress rate and the past progress rate at the same time. In these comparisons, numerical information is compared for each of the same items.

Among these comparisons, if the acquired company information is for the full-year financial results, a comparison between the newly announced earnings forecast for the current term and past results, and a comparison between the announced results for the previous term and the forecast values, The financial results for each three months will be compared with the past actual values, and the announced profit margin will be compared with the past profit margin. In addition, in such comparisons related to full-year financial results, in the comparison of financial results and past results, it is always compared whether the record high or the record high is expected. Among the numerical information (financial results) included in the company information, the past extreme values (record high or record low) will be compared with the newly announced financial results.

In addition, if the acquired company information is quarterly financial results and the full-year forecasts regarding business performance and dividends have been revised, the revised forecasted business performance will be compared with the previous forecast, and the revised forecasted dividend will be compared with the previous forecast. On the other hand, if no corrections have been made, comparison of the progress rate of the actual figures against the forecast, comparison of the progress rate with the settlement of accounts for the same period in the past, and comparison of the settlement figures of 3 months with the past actual figures. , The profit margin and the past profit margin are compared. Even in such a comparison of quarterly financial results, it is not possible to compare whether the financial results are set to a record high (or record low) or are expected to reach a record high (or record low). This is always done, and in this case as well, among the numerical information (financial results) included in the past company information, the past extremum (record high or record low) was newly announced. The settlement values will be compared.

Furthermore, at the stage of S4, analysis such as whether or not corrections are included in the announced company information is also performed by the article generator 20 (MPU20a), and among the items included in the announced company information. The article generator 20 (MPU20a) also performs a process of analyzing what the item affecting the stock price (referred to as the stock price driver) is. All financial items (sales, net income, EBITDA (interest rate / tax / profit before amortization), net assets, dividends, etc.) included in the announced corporate information are candidates for potential stock price drivers. By comparing the correlation between the numerical value related to the item of and the stock price (including the stock price distributed from the stock price distribution system 3) with the average value of other companies sent from the information DB system 10, the article generator 20 (MPU20a) ) Performs the process of identifying the stock price driver. The stock price driver affects the stock price and is of concern to investors.

Then, at the stage of S5 of the flowchart shown in FIG. 6, the article generation device 20 (MPU20a) performs a process of specifying the headline or the subject of the article based on the processing content of S4 described above. That is, in the processing of S4, the article generator 20 (MPU20a) uses the latest company information (settlement, etc.) announced in each received company information as the past company information (settlement, etc.) and Each item is compared with the earnings forecast information (forecast related to financial results, etc.), and the process of specifying the headline or subject content of the automatically generated article from the comparison result will be performed at the stage of S5.

The comparison, which is the premise of this specific processing, is to compare a plurality of numerical values (numerical information) related to the business performance of the company included in each information at the stage of S4 of the flowchart for each item. The article generator 20 (MPU 20a) calculates the points defined in the evaluation standard table 25 based on the results. Then, the article generator 20 (MPU20a) identifies the item related to the numerical information in which the situation corresponding to the difference between the numerical information is the largest, and specifies the content (information) corresponding to the item as a headline or a subject. Then, the specified content will be described in the heading or subject. In this embodiment, the calculated points (corresponding to the evaluation value according to the difference) are used as the situation according to the difference between the numerical information, and the item to be compared in which the calculated point value is the largest is specified. Then, the content (information) corresponding to the item will be described in the heading or subject.

As described above, the company information acquired by the article generator 20 (MPU 20a) from the information DB system 10 is the information indicated in a predetermined standardized file format (for example, XBRL format) (however, the numerical information is as described above. In such a standardized file format, multiple items representing achievements are listed in a predetermined order with the items and numerical values set. .. Therefore, if the announced company information is the settlement of accounts, the numerical value indicated by the numerical information corresponding to each item included in the newly announced settlement of accounts and the numerical value corresponding to each item included in the past settlement of accounts, In addition, the numerical values indicated by the numerical information corresponding to each item included in the forecasted settlement of accounts will be compared in order. Items included in the settlement of accounts include sales, operating income, profit before tax, ordinary income, total assets, dividends, etc., so the figures for each of these items will be compared in sequence.

Further, regarding the comparison result, points (evaluation values) are calculated based on the rules in the evaluation standard tables 25 (each table unit 25a, 25b) shown in FIGS. The article generator 20 (MPU20a) identifies the item to be compared with the largest equivalent), and determines to use the information corresponding to the specified item as the headline or means of the article (S5). As described above, in the present embodiment, since the heading or means is determined by identifying the point (evaluation value) having a large absolute value, the ordinary profit of the latest announced settlement of accounts is the ordinary profit of the settlement of accounts for the past. If the evaluation value according to the difference (difference in absolute value) is the largest among multiple items, not only in the case of more positive, but also in the case of negative, the content such as "significant reduction in ordinary income" is given. Will be used for headings or subjects.

In addition, it is an example to use the absolute value as the situation (point which is the evaluation value) according to the difference, and depending on the specifications of the system etc., the evaluation value (point) according to the difference is not used. The item may be specified only when is positive, and in such a case, it is suitable when it is desired to convey the largest positive item of the company to the user in an easy-to-understand manner. On the other hand, contrary to the above, it is also possible to specify the item only when the evaluation value (point) according to the difference is negative, and in this case, the largest negative item (minus) of the company. It is suitable for making the item with a large value) stand out and telling the user.

Then, at the stage of S6 of the flowchart shown in FIG. 6, the article generation device 20 (MPU 20a) performs the process of determining the content and order of the paragraphs constituting the article. In identifying the content of the paragraph, the ordinary income is basically the most important judgment criterion item (important item) among the multiple items included in the company information, so the item of the ordinary income is included in the paragraph content. Is always included. In addition, when the ordinary income included in the announced corporate information is in the black, but the net income is in the red, etc., the state of the ordinary income and the state of other profits are opposite. Then, since items other than ordinary income can be factors that affect the stock price, items that can be factors that affect such stock prices are treated as important items, and information related to those important items is also included in the contents of the paragraph. Decide that.

Further, the above-mentioned processing related to the stock price driver is also used for determining the content and order of paragraphs. That is, the numerical values (numerical information) corresponding to a plurality of items included in the acquired new company information are compared with the average numerical information included in the acquired information DB system 10, and both numerical values are based on the comparison result. Points indicating the situation according to the difference in information (ratio according to the difference: equivalent to the evaluation value of the difference) are calculated according to the evaluation standard table 25, and the points of the items related to the calculated points become larger. The order is specified, and the article is automatically generated by describing the information related to each item according to the order of the specified items. At this time, by dividing the description of the information related to each item into paragraphs for each description content, the content and order of the paragraphs are determined.

Then, at the stage of S7 in the flowchart shown in FIG. 6, the article generation device 20 (MPU20a) automatically generates an article of company information based on the contents of the above processes S4, 5, 6, and the like. Specifically, regarding the headline of the article, the headline sentence that describes the information corresponding to the item specified in the stage of S5 is added to the article by the function corresponding to the sentence creation engine included in the article generation program P2 (function by AI). Created by the generator 20 (MPU 20a).

Regarding the content of the article, the article generator 20 (MPU20a) first includes the subject specified in the stage of S5 and the important items determined in the stage of S6 by the function corresponding to the sentence creation engine. In addition to creating the paragraphs of, the sentences of the second and subsequent paragraphs are also created by including the information corresponding to the items determined in the stage of S6 in each paragraph.

Such sentence creation is basically a set of the above-mentioned items and numerical information in the XBRL format among the company information acquired from the information DB system 10. , Items are recognized as the subject, numerical information is recognized as the object, particles, verbs, etc. that connect the subject item and the numerical information of the object are determined by analysis by the sentence creation engine, and according to each paragraph. I will create concrete sentences. That is, for the words, numerical values, etc. corresponding to the subject and predicate used in each paragraph, the item contents and numerical information corresponding to the ordinary profit, important items, etc. that were processed at the stage of S6, etc. described above are used, and S3. Information on the earnings forecast included in the company information acquired at the stage of (information stored in the earnings forecast DB17), and the content of the result calculated by calculation at the stage of S4 (calculated items and numerical values of the calculated result), etc. You will use it to create sentences.

Regarding the first sentence of the first paragraph, the company name, the date and time of the announcement, and the type of announcement (settlement, revision of earnings forecast, revision of dividend, etc.) included in the newly announced company information are extracted. Then, the particles, adverbs, verbs, etc. that connect these extracted words are determined by analysis by the sentence creation engine to create the first sentence of the first paragraph. Then, the company information articles created in this way are separately stored by the MPU 20a in the created article table 24 (see FIG. 3) included in the storage unit 20g for each company corresponding to the articles.

Finally, at the stage S8 of the flowchart shown in FIG. 6, the article generator 20 (MPU20a) displays the article (company information article) generated at the stage S7 above with the user terminals T1, T2, T3 shown in FIG. Deliver to etc. via the network. This delivery can be assumed to be sent to the target user terminal as a push delivery, or to be sent to the requesting user terminal in response to a request (request) from the user terminal side, and is used for commercial purposes. , You may charge for the delivery. In addition, the distribution of the company information article at the stage of S8 is performed instantly (in seconds within 10 seconds) or within 1 minute from the stage of S1 at the earliest, and is usually about 2 to 3 minutes in minutes. Even if many presentations are made at the same time and multiple articles are generated in parallel, it will be done within 4 to 6 minutes at the latest.

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. Similar to the article K10 shown in FIG. 14A, this company information article 30 is included in the financial statements (corporate information: financial statements) of company A disclosed (announced) by the stock exchange system 2 of the Tokyo Stock Exchange and the like. Based on this, it is automatically generated by the present invention.

The company information article 30 has a heading section 30a, a first paragraph section 30b, a second paragraph section 30c, a third paragraph section 30d, and a delivery time section 30e. The heading section 30a indicates that Company A's ordinary profit for the May-January period increased by 51%, and that the ordinary income for the 11-January period also increased 2.3 times. Is specified at the stage of S5 in the flowchart of FIG. As shown in FIG. 7, the content of the heading portion 30a clearly shows the status of the ordinary profit of the company A, and the article K10 by the same conventional technique of the company A shown in FIG. 14 (a) is uniform. It is different from the content of the typical headline.

In addition, the first paragraph part 30b of the company information article 30 first contains a sentence indicating the company name (company A), the date and time of the announcement (15:00 on March 1), and the type of announcement (settlement), and then. , Contents of consolidated ordinary income for the cumulative third quarter, which corresponds to the subject and important matters, compared with the same period of the previous year (up 51.4% from the same period of the previous year to 17.5 billion yen), progress rate against the full-year plan (forecast) , And comparisons with the past 5 years. Such a first paragraph part 30b makes it easy for the readers of the article to understand the situation regarding ordinary income, and also shows the progress rate for the plan, so that the article accurately conveys the information of the company that the investor is watching. It is the content. The progress rate in the first paragraph 30b and the comparison result with the 5-year average use the processing contents at the stage S4 of the flowchart shown in FIG. On the other hand, in the article K10 based on the prior art in FIG. 14 (a), the contents of the four items of net profit, sales, ordinary profit, and operating profit are listed in the first paragraph. It is difficult to convey to the readers of the article whether or not is important. Therefore, the content related to ordinary profit, which was prominently described in the article information article 30 of the present invention, is also enumerated in the article K10. It's buried inside.

Furthermore, the second paragraph 30c of the corporate information article 30 compares the forecast value of consolidated ordinary income for the February-April period with the same period of the previous year, and the content of the relevant company that investors care about next fiscal year. The situation is shown as a forecast. The contents of the forecast status for the next fiscal year are based on the comparison between the actual results for the previous fiscal year and the forecasts for the next fiscal year specified in the evaluation standard table 25 (stage S4 in the flowchart of FIG. 6). On the other hand, in the article K10 based on the prior art shown in FIG. 14 (a), the forecast for the next fiscal year is described in the second paragraph, which also includes net income, sales, ordinary income, as in the first paragraph. And because the forecast contents of the four items of operating income are listed, it is difficult for the reader of the article to know which item is important.

Furthermore, the third paragraph part 30d of the corporate information article 30 has the contents of comparing the consolidated ordinary income and the operating income margin, which are the results of the last three months, with the same period of the previous year. The content is useful information for investors, etc. to grasp the latest performance results of Company A. On the other hand, the article K10 according to the prior art of FIG. 14A does not include the content corresponding to the third paragraph part 30d of the corporate information article 30.

Summarizing the above, if an investor or the like reads the company information article 30 regarding company A shown in FIG. 7 according to the present invention, the current performance of company A can be found from the contents of the heading section 30a and the first paragraph section 30b. It is going well, and this is supported by the contents of the third paragraph part 30c, but from the contents of the second paragraph part 30c, it is judged that the current situation is the peak state of the business performance because the business performance will be on the decline in the future. Although the content conveys the materials that can be produced in an easy-to-understand manner, it is difficult to read the above content that can be judged from the company information article 30 of FIG. 7 from the article K10 by the conventional technique of FIG. 14 (a). You can see that there is.

Further, FIG. 8 shows a company information article 31, which is another example of an article automatically generated by the processing of the flowchart shown in FIG. 6, and the company information article 31 is the same as the article K11 shown in FIG. 14 (b). It is automatically generated based on the financial statements (company information: financial statements) of Company B disclosed (announced) by the stock exchange system 2 of the Tokyo Stock Exchange.

Similar to the company information article 30 shown in FIG. 7, the company information article 31 has a heading section 31a, a first paragraph section 31b, a second paragraph section 31c, a third paragraph section 31d, and a delivery time section 31e. The heading section 31a shows that Company B's ordinary profit for the current term increased by 3%, the highest profit for the fifth consecutive term, and that the dividend will increase by 11 yen. It is the one specified at the stage of S5 in the flowchart of 6. The content of the heading section 31a is because the content of the heading of the article K11 by the same prior art of the same company B shown in FIG. 14B is related to net profit (Company B's net profit for the fiscal year ended January 31, 2017). (44.5% increase of 121.5 billion yen), in the company information article 31 about Company B, the heading contents of both articles are different as in the case of both articles shown in FIGS. 7 and 14 (a). ..

The first paragraph 31b of this company information article 31 is the subject or important after the sentence indicating the company name (company B), the date and time of the announcement (15:30 on March 9), and the type of announcement (financial statements). As important items, the consolidated ordinary income for the fiscal year ending January 2017 includes the content compared with the previous fiscal year (up 18.9% from the previous fiscal year to 190.9 billion yen), and the ordinary income for the fiscal year ending January 31, 2018 is also compared with the previous fiscal year. On the other hand, since it has sentences including contents that are expected to grow and record high profits for the fifth consecutive term, it is easy to tell the readers of the article that Company B's business performance is good and it is likely to continue. It is the content. The contents of the first paragraph section 31b (current state, future contents, record high, etc. regarding ordinary profit) were obtained by the processing contents in the stages S4 to 6 of the flowchart shown in FIG.

The content of the first paragraph section 31b of the corporate information article 31 is different from the content of the first paragraph section 30b of the performance information article 30 shown in FIG. 7 above in that it includes future performance prospects. Therefore, in the present invention, it can be seen that articles having individual contents are automatically generated according to the performance of each company. In addition, in the article K11 by the conventional technology shown in FIG. 14 (b), which is the same article by Company B, in the first paragraph, the contents of four items of net profit, sales, ordinary profit, and operating profit are listed. Since it is limited to the standard expression described, it is difficult for the reader of the article to know which item is important, as in the article K11 of FIG. 14 (a).

Furthermore, the second paragraph section 31c of the corporate information article 31 is intended to increase the annual dividend for this term, and conveys the content of the dividend that investors care about. It should be noted that the content related to such a dividend is obtained by the processing content of the stage such as S4 in the flowchart of FIG. The second paragraph 31c of the company information article 31 is also different from the content of the second paragraph 30c of the company information article 30 shown in FIG. 7, and again, individual article contents are generated for each company. You can see that there is. In addition, in the article K11 by the prior art shown in FIG. 14 (b), which is the same article of the company B, the forecast for the next fiscal year is described in the second paragraph, so that the second paragraph portion 31c of the company information article 31 The content is completely different from.

Furthermore, the third paragraph section 31d of the corporate information article 31 shows the contents of the consolidated ordinary income and the operating profit margin of the last three months compared with the same period of the previous year, and the contents of the third paragraph section 31d. Is useful information for investors and others to grasp the latest business performance information of Company B. On the other hand, the article K11 according to the prior art of FIG. 14B does not include the content corresponding to the third paragraph portion 31d of the corporate information article 31.

Summarizing the above with respect to the company information article 31 of FIG. 8, if an investor or the like reads the company information article 31 relating to company B shown in FIG. 8 according to the present invention, the heading section 31a and the first to third paragraph sections From the contents of 31b to 31d, the current performance of Company A is favorable, and the content clearly conveys the material that can be judged that the good performance is likely to continue for a while. ) From the article K11 based on the prior art, it can be seen that it is difficult to smoothly grasp the above contents as can be read from the company information article 31 of FIG.

FIG. 9A shows a company information article 40 which is an example of an article automatically generated by the processing of the flowchart shown in FIG. 6, and the company information article 40 is provided by a stock exchange system 2 such as the Tokyo Stock Exchange. It is automatically generated based on the disclosed (announced) material of Company C (company information: revised earnings forecast).

The company information article 40 has a heading section 40a, a first paragraph section 40b, a second paragraph section 40c, and a delivery time section 40d. The heading section 40a indicates that Company C has revised up the ordinary profit for the current term by 16%, and this content was specified at the stage of S5 in the flowchart of FIG. In addition, the first paragraph 40b of Corporate Information Article 40 indicates that the consolidated ordinary income for the fiscal year ending March 31, 2017 has been revised upward from the previous forecast as a subject or important item, and the profit decline rate is expected to decrease. At the same time, the consolidated final profit for the full year has been revised downward from the previous forecast, and the profit decline rate is expected to increase, so the content of the article accurately conveys the business performance trends of companies that investors care about. It should be noted that the content of the first paragraph portion 40b is also based on the processing content in the steps S4 to S6 of the flowchart shown in FIG.

Furthermore, the second paragraph 40c of the corporate information article 40 shows that the consolidated ordinary income and the profit growth rate in the second half will increase due to this revision announcement, and this content is also shown in FIG. It is based on the processing contents in the stages S4 to S6 of the flowchart. As described above, in the present invention, when a company announces a revision of its earnings forecast, an article is automatically generated to inform investors of the content that affects the stock price, while the above-mentioned non-patent document of the prior art Since the article is generated in 1 only when the company announces the financial statements, the article cannot be generated when the earnings forecast revision is announced in this way.

FIG. 9B also shows an example company information article 41 automatically generated by the processing of the flowchart shown in FIG. 6, and this company information article 41 is disclosed (announced) by a stock exchange system 2 such as the Tokyo Stock Exchange. ) Is automatically generated based on the materials of Company D (company information: revised earnings forecast).

The corporate information article 41 also has a heading section 41a, a first paragraph section 41b, a second paragraph section 41c, and a delivery time section 41d. The heading section 41a indicates that Company D will raise the ordinary profit for the current term by 16% to renew the highest profit, and this content was also specified at the stage of S5 etc. in the flowchart of FIG. It has become. In addition, in the first paragraph section 41b of Corporate Information Article 41, as a subject or important item, the consolidated ordinary income for the fiscal year ending March 2017 has been revised upward from the previous forecast, and it is expected to increase by 16%. The text indicates that it is expected to record a record high profit for the consecutive term. The second paragraph, 40c, indicates that the consolidated ordinary income and the profit growth rate in the second half will be increased and expanded, respectively, due to the announcement of this revision. This content is also shown in S4 of the flowchart shown in FIG. It is obtained by the treatment in the steps of ~ S6.

FIG. 10 shows a company information article 42 which is an example of an article automatically generated by the processing of the flowchart shown in FIG. 6, and this company information article 42 is also disclosed (announced) by a stock exchange system 2 such as the Tokyo Stock Exchange. ) Is automatically generated based on the materials of Company E (company information: dividend correction).

The corporate information article 42 has a heading section 42a, a first paragraph section 42b, and a delivery time section 42c. The heading section 42a states that Company E will return the dividend for the current term at 20 yen for the first time in two terms, and this content is specified at the stage such as S5 in the flowchart of FIG. In addition, the first paragraph section 42b of the corporate information article 42 indicates that the undecided dividend for the fiscal year ending March 2017 will be paid and the dividend will be returned for the first time in two terms, so that the situation of the company that investors care about is accurate. It is the content of the article to tell to. It should be noted that the content of the first paragraph portion 42b is also based on the processing content in the steps S4 to S6 of the flowchart shown in FIG. As described above, in the present invention, even if the company announces the dividend adjustment, the article is automatically generated, while the article is generated in the above-mentioned non-patent document 1 only when the company announces the financial statements. Therefore, if the dividend revision is announced in this way, the article cannot be generated.

As described above, in the present invention, when each company announces company information such as financial results, an article focusing on notable matters is automatically created according to the content of the announcement instead of the standard content. , It has the merit of being able to quickly provide investors with articles that express matters that must be considered and judged in investment, etc. in an easy-to-understand manner, and articles with such contents are business results other than financial results. The feature is that it is generated even when forecast revisions, dividend revisions, etc. are announced.

The present invention is not limited to the above-described embodiment, and various modifications can be assumed. For example, in the above explanation, the average value stored in the past settlement DB 16 is the average value of the remaining other companies excluding one company, but the other companies are one company. It is also possible to narrow down to companies in the same industry, calculate an average value for each industry, and perform comparisons according to the industry. In addition, when trying to simplify the specifications and reduce the burden of the average value calculation process, calculate not the average value excluding one company, but multiple average values including one company without excluding it. The average value calculated in this way may be used for comparison or the like.

Further, although the generated article is provided with a heading portion 30a or the like indicating the heading of the article, the specification may omit the provision of the heading portion 30a or the like depending on the specifications of the article to be generated. In this case, the content specified in the stage S5 of the flowchart of FIG. 6 should be described in the subject of the article, and the subject of the article should be the company name and the text of the announcement time in the first paragraph part 30b or the like. Include it in the following text to make the subject stand out.

Further, in the above description, the headline or subject of the article is the numerical information contained in the newly announced corporate information, both the numerical information contained in the performance forecast information and the numerical information contained in the past corporate information. Although it was designed to be specified by comparison, if the specifications are to be simplified, it may be specified by comparing only one of them (included in the numerical information and performance forecast information included in the new company information). It is also possible to specify the headline, etc. by either comparing the numerical information contained in the new corporate information or the numerical information contained in the new corporate information with the numerical information contained in the past corporate information).

In addition, although the article generated in the above processing did not include the content related to the stock price, the information DB system 10 provides information on the stock price (theoretical stock price) stored in the theoretical stock price DB18 and the individual investor stock price forecast DB19. , Expected stock prices by a plurality of individual investors, etc.) are also sent to the article generator 20, and the article generator 20 also receives the latest stock price (actual stock price) of each company distributed by the stock price distribution system 3. Since it is acquired, information on these stock prices may be included in the generated article.

FIG. 11A shows an example of the stock price information 50 of the company A including the latest stock price 50a (actual stock price) of the company A and the theoretical stock price 50b of the company A, and the stock price information 50 of the company A is shown in, for example, FIG. It may be arranged in an article such as below the heading portion 30a of the company information article 30 shown in 7 or above the delivery time portion 30e to provide the content related to the stock price together with the article. In this case, as the stock price 50a of company A included in the stock price information 50 of company A, the latest stock price acquired from the stock price distribution system 3 is used when the article is generated, and the theoretical stock price 50b of company A is , The value of the theoretical stock price of company A read from the theoretical stock price DB 18 included in the information acquired at the stage S3 of the flowchart of FIG. 6 is used.

Further, FIG. 11B shows an example of the stock price information 51 of the company A including the stock price 51a (actual stock price) of the company A and the expected stock price 51b of the company A, and the stock price information 51 of the company A is shown in FIG. Similar to the case of the stock price information 50 of company A in (a), for example, it is arranged in an article such as below the heading section 30a of the company information article 30 shown in FIG. 7 or above the delivery time section 30e, and together with the article. Content regarding stock prices may be provided. In this case, as the stock price 51a of company A included in the stock price information 51 of company A, the latest stock price of company A is used as in the case shown in FIG. 11 (a), and the expected stock price 51b of company A is , The value of the forecasted stock price of Company A read from the individual investor stock price forecast DB 19 included in the information acquired at the stage S3 of the flowchart of FIG. 6 is used.

As another modification, it is also possible to use stock price information such as a combination of FIGS. 11A and 11B. In this case, for example, the stock price of company A (latest stock price), A. Stock price information including the theoretical stock price of company A and the expected stock price of company A will be posted together with the article. By posting the stock price information shown in FIGS. 11 (a) and 11 (b) together with the article, it becomes possible to confirm the influence and relevance of the stock price and the content of the article together.

Furthermore, in order to make it possible to understand at a glance the outline of the announcement that caused the article to be generated in the articles related to the automatically generated financial statements (corporate information articles 30, 3140 to 42, etc. shown in FIGS. 7 to 10 etc.). The overall evaluation of the announcement (overall evaluation of the announcement regarding financial results, etc.) may be shown together with the article.

FIG. 11 (c) shows an example of a comprehensive evaluation 52 (an example of an overall evaluation) for the announced financial results of Company A, and in this example, five stars so that the contents of the financial results can be evaluated on a five-point scale. It is designed to fill the contents of the star symbol according to the 5-grade evaluation (for example, if the evaluation is 3 out of 5, the star symbol from the left to the third is displayed. Will be filled). In such a five-step evaluation, the points of each item calculated based on the evaluation standard table 25 are totaled, and the five-step evaluation is performed according to the value of the total points. In such an evaluation, points are calculated at the stage of S4 in the flowchart of FIG. 6, the value of the total points is obtained, and the stage of evaluation is determined based on the value of the total points. For example, if the value of the obtained total points is 5, the evaluation will be 5 out of 5 levels, and similarly, if the value of the total points is 4, the evaluation will be 4 out of 5 levels, the value of the total points. If is 3, the evaluation is 3 out of 5 grades, if the total point value is 2, the evaluation is 2 out of 5 grades, and if the total point value is 1, the evaluation is 1 out of 5 grades. Become.

The comprehensive evaluation 52 of company A generated based on the determined evaluation is placed below the heading section 30a of the company information article 30 shown in FIG. 7, or above the delivery time section 30e, and is announced together with the article. Provide an evaluation of financial results. By providing the comprehensive evaluation 52 together with the article in this way, investors and the like can grasp whether the financial results announced at a glance are favorable or unfavorable by the comprehensive evaluation 52, and then more detailed information. It is possible to create a flow of checking the content in an article, which can contribute to making smooth decisions by investors.

As a method of evaluation, in addition to the above-mentioned five-grade evaluation, it is conceivable to increase the number of grades and evaluate with six or more grades, or conversely reduce the number of grades and evaluate with three or less grades. Furthermore, it is also possible to evaluate as an overall evaluation in two stages of plus or minus.

In this way, even when the overall evaluation of the announced company information is performed in two stages such as plus or minus, the numerical information of each item included in the new company information is included in each of the performance forecast information to be compared. It is compared with the numerical information of each item or the numerical information of each item included in the past company information, and in the result of the comparison, it is calculated for each item according to the conditions specified in the evaluation standard table 25 as described above. It will be done based on the points. However, when the calculated points for multiple items are mixed with plus and minus, that is, the numerical information of each item included in the newly announced company information becomes larger than the numerical information to be compared. When the case and the case become smaller, it is difficult to make a uniform evaluation as a whole, so in such a case, it is decided to make an overall evaluation based on the important items among multiple items. Become.

That is, when the numerical information included in the new company information is larger than the numerical information to be compared in the comparison between the numerical information in the important items (when the points calculated in the important items are positive), A positive overall evaluation (comprehensive evaluation) will be conducted for the newly announced company information. On the other hand, when the numerical information included in the new company information is smaller than the numerical information to be compared (when the calculated points are negative in the important items), a negative overall evaluation (comprehensive evaluation) is performed. It will be. Then, if such a positive or negative overall evaluation is attached to an automatically generated article (company information article 30 etc.) and distributed to users such as investors, the investor etc. will receive the announced company information. , You can quickly judge whether it is positive or negative in general by the overall evaluation attached to the article.

Ordinary income is basically used as an important item among multiple items (items related to ordinary income are made important items). However, when ordinary profit is in the black but net profit is in the red, and there are both surplus and deficit in the items related to profit, items other than ordinary profit become items that affect the business performance of the company. In such a case, items that affect business performance other than ordinary income will be important items. To identify specific important items in such cases (when specifying items other than ordinary income as important items), the average numerical value is the numerical information corresponding to multiple items included in the newly announced corporate information described above. When comparing information and each item, the article generator 20 identifies the item to be compared as an important item, which has the largest value (point) indicating the situation according to the difference between the numerical information and the average numerical information. ..

Further, in the above explanation, the generated articles (corporate information articles 30, 31, 40 to 42, etc. shown in FIGS. 7 to 10 and the like) are text-based, but the content of the articles is based on graphs and symbols. It is also possible to visualize.

FIG. 12 shows a visualized example of corporate information article 60 (article for announced financial results), and includes a summary unit 61, a graph unit 62, and an evaluation unit 63 in order from the top. In this modified example, in order to visualize (design) the article, the settlement contents are narrowed down to the items of "sales", "operating profit", and "final profit". For these three items, a symbol indicating the evaluation of "Good!" Or "Bad" is attached together with the summary. Further, in the graph unit 62, these three items are represented by a graph with each period on the horizontal axis. The evaluation unit 63 is basically the same as that shown in FIG. 11 (c). Since the content of such a visualized corporate information article 60 can be visually grasped as compared with a text article, it is suitable for investors and the like who desire to understand corporate information speedily.

Since the present invention automatically generates an article edited so that important matters included in corporate information such as announced financial statements are the main items, it is used to provide matters that investors and others care about in an easy-to-understand expression. Can be suitably used for.

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 settlement database
17 Earnings forecast DB
18 Theoretical stock price DB
19 Individual Investor Stock Price Forecast DB
20 Article generator 20a MPU
20g Storage unit 24 Created article table 25 Evaluation criteria table P2 Article generation program

Claims (15)

In an article generation system equipped with 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 includes numerical information corresponding to multiple items related to the company's business performance.
It is equipped with an information database system that receives company information distributed from the information distribution source and stores forecast company information including numerical information of forecasts according to the plurality of items.
The article generator is
A means for detecting that new company information distributed from an information distribution source is received by the information database system, and
When the reception of new company information by the information database system is detected, the means for acquiring the new company information and the predicted company information of the same company as the company related to the new company information from the information database system.
A means to compare the numerical information according to multiple items included in the acquired new company information with the numerical information of the forecast according to multiple items included in the acquired forecast company information, and a means for comparing each item.
Among the compared numerical information, the means for identifying the item to be compared, which has the largest difference between the numerical information, and
An article generation system characterized in that it includes a means for describing information according to a specified item in the headline or subject of a company information article. The information database system is
The company information received from the information distribution source is stored as past company information.
The article generator is
When the reception of new company information by the information database system is detected, in addition to the new company information and the predicted company information, the past company information of the same company as the company related to the new company information is added to the information. Means to get from the database system
Numerical information according to a plurality of items included in the new company information, according to a plurality of items included in the numerical information of the forecast according to the plurality of items included in the forecast company information and the acquired past company information. The article generation system according to claim 1, further comprising a means for comparing past numerical information with each item. The information database system stores average numerical information obtained by averaging numerical information corresponding to a plurality of items included in corporate information of a plurality of companies.
The article generator is
A means of acquiring average numerical information in line with the acquisition of new company information,
A means to compare the numerical information corresponding to multiple items included in the acquired new company information with the acquired average numerical information for each item,
Based on the comparison, a means to specify the order of the items to be compared, in which the situation of the difference between the numerical information and the average numerical information becomes large,
The article generation system according to claim 1 or 2, further comprising a means for automatically generating a company information article by describing information related to each item according to the order of the specified items. The information database system includes means for storing the theoretical stock price theoretically calculated for each company's stock price.
The article generator is
A means of acquiring the theoretical stock price of a company according to the new company information in accordance with the acquisition of new company information,
How to get the actual stock price of each company to be delivered,
The article generation system according to any one of claims 1 to 3, further comprising a means for automatically generating a company information article including the acquired theoretical stock price and the actual stock price of the same company as the company related to the theoretical stock price. .. The information database system is provided with a means for storing the investor's expected stock price calculated by aggregating the stock prices predicted by a plurality of investors for the stock price of each company.
The article generator is
A means of acquiring the investor's expected stock price of a company according to the new company information in accordance with the acquisition of new company information,
How to get the actual stock price of each company to be delivered,
The description in any one of claims 1 to 4, which comprises a means for automatically generating a company information article including the acquired investor's expected stock price and the actual stock price of the same company as the company related to the investor's expected stock price. Article generation system. In an article generator that automatically generates company information articles based on company information including numerical information corresponding to multiple items related to the performance of the company distributed from the information distribution source.
The fact that new company information is received by an external information database system that receives company information delivered from the information distribution source and stores forecast company information including numerical information of forecasts according to the plurality of items. Means to detect and
When the reception of new company information by the external information database system is detected, the new company information and the predicted company information of the same company as the company related to the new company information can be obtained from the external information database system.
A means to compare the numerical information according to multiple items included in the acquired new company information with the numerical information of the forecast according to multiple items included in the acquired forecast company information, and a means for comparing each item.
Among the compared numerical information, the means for identifying the item to be compared, which has the largest difference between the numerical information, and
An article generation device comprising a means for describing information according to a specified item in a headline or subject of a company information article. The external information database system accumulates the company information received from the information distribution source as the past company information.
When the reception of new company information by the external information database system is detected, in addition to the new company information and the forecast company information, the past company information of the same company as the company related to the new company information is externally obtained. Means to obtain from the information database system of
Numerical information according to a plurality of items included in the new company information, according to a plurality of items included in the numerical information of the forecast according to the plurality of items included in the forecast company information and the acquired past company information. The article generator according to claim 6, further comprising past numerical information and a comparison means for comparing each item. The comparison means according to claim 7, wherein the numerical information included in the new company information is compared with the numerical information which is the past extreme value in the numerical information included in the past company information. Described article generator. The plurality of items include important items.
When the numerical information included in the new company information includes both large and small cases with respect to the numerical information to be compared, the numerical value included in the new company information is compared by comparing the numerical information related to important items. If the information is larger than the numerical information to be compared, the evaluation means for positive overall evaluation of the new company information and the evaluation means.
The article generator according to any one of claims 6 to 8, further comprising a means for attaching the positive overall evaluation to a company information article. When the numerical information included in the new company information includes both a large case and a small value with respect to the numerical information to be compared, the evaluation means is included in the new company information in the comparison relating to the important item. If the numerical information to be compared is smaller than the numerical information to be compared, the new company information is subjected to a negative overall evaluation.
The article generator according to claim 9, further comprising a means for attaching the negative overall evaluation to a company information article. The article generating device according to claim 9 or 10, wherein the important item is an ordinary profit. The external information database system stores average numerical information obtained by averaging numerical information corresponding to multiple items included in the corporate information of multiple companies.
A means of acquiring average numerical information from an external information database system in accordance with the acquisition of new company information,
A means to compare the numerical information corresponding to multiple items included in the acquired new company information with the acquired average numerical information for each item,
The article generator according to claim 9 or 10, further comprising means for setting the item to be compared, which has the largest difference between the numerical information and the average numerical information, as the important item based on the comparison. The external information database system stores average numerical information obtained by averaging numerical information corresponding to multiple items included in the corporate information of multiple companies.
A means of acquiring average numerical information from an external information database system in accordance with the acquisition of new company information,
A means to compare the numerical information corresponding to multiple items included in the acquired new company information with the acquired average numerical information for each item,
Based on the comparison, a means to specify the order of the items to be compared, in which the situation of the difference between the numerical information and the average numerical information becomes large,
The article generating device according to claim 6 to 12, further comprising a means for describing information related to each item and generating a company information article according to the order of the specified items. In an article generation method in which an article generator automatically generates a company information article based on company information including numerical information corresponding to a plurality of items related to the performance of the company distributed from the information distribution source.
The article generator is
The fact that new company information is received by an external information database system that receives company information delivered from the information distribution source and stores forecast company information including numerical information of forecasts according to the plurality of items. Steps to detect and
When the reception of new company information by the external information database system is detected, the step of acquiring the new company information and the predicted company information of the same company as the company related to the new company information from the external information database system, and
A step to compare the numerical information according to multiple items included in the acquired new company information with the numerical information of the forecast according to multiple items included in the acquired forecast company information for each item.
Among the compared numerical information, the step to identify the item to be compared where the difference between the numerical information is the largest, and
An article generation method including a step of describing information according to a specified item in a headline or subject of a company information article. In a computer program for causing a computer to automatically generate a company information article based on company information including numerical information corresponding to a plurality of items related to the performance of the company distributed from the information distribution source.
On the computer
The fact that new company information is received by an external information database system that receives company information delivered from the information distribution source and stores forecast company information including numerical information of forecasts according to the plurality of items. Steps to detect and
When the reception of new company information by the external information database system is detected, the step of acquiring the new company information and the predicted company information of the same company as the company related to the new company information from the external information database system, and
A step to compare the numerical information according to multiple items included in the acquired new company information with the numerical information of the forecast according to multiple items included in the acquired forecast company information for each item.
Among the compared numerical information, the step to identify the item to be compared where the difference between the numerical information is the largest, and
A computer program characterized by performing a step of describing information according to a specified item in the headline or subject of a company information article.

Images