KR101020138B1 - Method and apparutus for automatic contents generation - Google Patents

Abstract

Disclosed are a method and apparatus for automatically generating content. According to an aspect of the present invention, there is provided a method for automatically generating content to receive information data from an information source system, and to store at least some of the received information data or past information data related to the information data. Generating processing data based on at least one of cumulative data DBs; and generating content based on at least one of the generated processing data, past processing data, the cumulative data DB, or the information data It includes a step.

Description

Method for generating content automatically and its device {Method and apparutus for automatic contents generation}

The present invention relates to a method for automatically generating content and a device thereof, and more particularly, to receive a breaking news article and not to simply deliver or organize the breaking news article, but to relate to the breaking news article based on past accumulated data and related information. The present invention relates to a method and an apparatus in which beneficial additional information, an analysis opinion, and the like can be automatically generated and provided to an information user.

The present invention provides a technical idea that can automatically generate content. In the present specification, content refers to information having a useful meaning to an information consumer, and may be used to include all information that may be generated according to the technical spirit of the present invention, regardless of form. For example, real estate quotes, product information, weather information, life insurance association, asset management association fund-related information may be the content. In addition, the content may be provided in any form such as text, image, table, graph, audio information and / or video information.

Hereinafter, for convenience of description, the contents will be described as an example of a case in which the contents are disclosure data and various useful information related to the stock, but the scope of the present invention is not limited thereto.

News articles from online newspapers that provide securities related articles have a high proportion of breaking news reports (straight articles) that must be reported urgently. However, in order to provide such a breaking news, it has been a lot of time, labor and cost because it is made manually by the hands of the journalist as shown in Figure 1a.

1A is a diagram for describing a conventional content generation method.

Referring to FIG. 1A, an information source system (eg, financial supervisory system, etc.) may be used. At around 15:26, performance-related disclosure information (1) was provided for Handok Pharmaceuticals (stock unique number: 002390). Then, the existing information providing service provider (eg, online newspapers) wrote a news article by hand based on the performance-related disclosure information (1) of Handok Pharmaceuticals (stock unique number: 002390). For example, as illustrated in FIG. 1A, the newspaper A produces a press article 2 manually based on the performance-related disclosure information 1 and provides the information consumer at about 15:29. In addition, the newspaper B produced the press article 3 by hand based on the same performance-related disclosure information 1 and provided it to the information consumer at about 15:35. In addition, the newspaper C also similarly prepared the press article 4 on the basis of the performance-related disclosure information 1 and provided it to the information consumer at about 16:22.

As can be seen in Figure 1a, such a conventional method has a disadvantage in that it takes time for writing an article by hand. That is, in the case of newspaper A, the reporting time is 15:29:59 seconds on October 23, 2006, in the case of newspaper B: 15:35:18 seconds on October 23, 2006, and in the case of newspaper C: 2006 At 23:22:17 on 23 October, it took 3 minutes, 2 seconds, 8 minutes, 21 seconds, and 55 minutes, 30 seconds, respectively. Furthermore, the news reporter (2), the news reporter (3), and the news reporter (4) all reported the original disclosure once more without additional data or analytical opinions. However, there is a problem that the quality of the information is not more than or less than the published original.

In addition, in the technical problem of the present invention, if you want to provide additional information, such as a variety of additional information other than the original disclosure to the information consumer, newspapers spend a lot of time manually generating such additional information to create securities content to compete There is a problem that the value as breaking news content is damaged.

Accordingly, when the technical problem of the present invention receives the performance-related disclosure information 1 illustrated in FIG. 1A, various additional information and / or analysis opinions related to the performance-related disclosure information 1 are illustrated as shown in FIG. 1B. Is automatically generated and quickly provided to the information consumer. For example, as shown in FIG. 1B, the technical problem of the present invention is a table comparing the past quarterly results of Handok Pharmaceuticals with the past results in relation to the above performance related disclosure information (1), and analysis of past stock price trends. On the basis of the information provided in the performance-related disclosure information (1), information such as ROE (operating profit margin), PER, and / or PBR, which are unknown in the performance-related disclosure information (1), and / or in the same industry You can automatically generate and provide content comparison information with other events as a content.

 On the other hand, there is a prior art patent No. 0416953, "System news providing system and method (hereinafter referred to as prior patent)". The preceding patent has a structure for generating news corresponding to logic preset by the 'user or system operator' with respect to the investment information. However, in the case of the preceding patent, since it is a structure that generates a fixed phrase for an event corresponding to a predetermined logic set by an information consumer or a system operator, information flowing out at all times is quantitatively and qualitatively enough to be called an article. Incompleteness

In addition, if only the information consumer is informed about the logic set by the information consumer, it is not the breaking news content but the real-time search function currently used in the home trading system (HTS). And the result it shows is fragmentary, and it is hard to say that it is an article that is meaningful only to some users who are trading stocks, such as the upper and lower limit information, and the selling amount information. In addition, since the base data, which is the basis of news generation, is also based on a single data such as quote information, the generated article may be regarded as a simple alarm device rather than a news. In addition, the preceding patent determines whether or not news occurs in a single data logic, which means that only 'information that the system operator or information consumer thinks meaningful' is generated as an article. . It is left to the arbitrary judgment of the system operator or the information consumer. It is difficult to say that the news is difficult because it can miss a large number of important articles and generate different news for each client. There is a problem. In addition, the preceding patent asks 'is there a case corresponding to logic?' Is an important determinant and content of news creation, so the content to be delivered is the occurrence of logic-related events.

As a result, the preceding patent does not provide a technical idea for solving the technical problem that the applicant intends to achieve.

Therefore, by processing complex and diverse disclosure information, it is possible to automatically generate press release contents to provide important information quickly, and filter the disclosure information even if there is a description error of disclosure information, such as a company's disclosure officer's input error. Therefore, only the valid information is generated as the content, which improves the reliability of the content, and the language processing necessary for the analysis of logic, sentence semantic parsing, and synthesizing the opinion or processing information to the extent that it can feel as if a person wrote the sentence. There is an urgent need for a method and an apparatus capable of providing complete content using technology.

Accordingly, the technical problem to be achieved by the present invention is to provide the information consumer by generating the information data itself in the form of a table or graph so that the information consumer can easily access the information data (for example, the disclosure information). The present invention provides a method and apparatus for automatically generating and providing various additional information and analysis opinions related to information data.

In addition, if it is determined that there is an error of the information data by checking an input error or the like of the received information data, the content is not generated, thereby providing reliable content.

According to an aspect of the present invention, there is provided a method for automatically generating content to receive information data from an information source system, and to store at least some of the received information data or past information data related to the information data. Generating processing data based on at least one of cumulative data DBs; and generating content based on at least one of the generated processing data, past processing data, the cumulative data DB, or the information data It includes a step.

The generating of the processed data includes parsing the received information data to extract at least a portion of the information data, and storing at least a portion of the extracted information data in the cumulative data DB, among the extracted information data. And generating the processed data based on at least one of at least one of data stored in the cumulative data DB or real-time data corresponding to the information data.

Receiving the information data may include receiving the information data corresponding to the search word from the information provider system based on a pre-stored search word.

The generating of the processed data may further include checking validity of the received information data. If the processed data is valid, at least a part of the information data may be extracted from the accumulated data DB and stored in the accumulated data DB.

The apparatus for automatically generating content may extract at least a portion of the information data or check the validity of the content data based on previously stored form information.

The generating of the content may include determining content data that is a part to be generated as the content from at least one of the generated processing data, past processing data, the cumulative data DB, or the information data, and determining the determined content data. And converting it into at least one of text data, a table, or a graph.

The converting of the determined content data into at least one of text data, a table, or a graph may include extracting a keyword from the determined content data, determining a term corresponding to the extracted keyword, and a numerical value corresponding to the keyword. If there is data, the method may include determining numerical data, rearranging the term, determined numerical data, and adding a sentence component based on the rearrangement result.

The converting of the determined content data into at least one of text data, a table, and a graph may include extracting a keyword included in the content data, comparing numerical data corresponding to the keyword, and a term corresponding to a comparison result. Selecting and determining analysis information based on the selected term and the keyword, determining the sentence component to be included between each of the determined analysis information, the keyword, or the term, and determining the determined sentence component, the analysis information, the Combining the keyword and the term into text data.

The automatic content generation method may further include converting at least a portion of the generated content into a foreign language. The processed data may include at least one of a stock price return ratio, an operating profit margin, or a return on equity, and the information data may be a stock-related breaking news article.

The generating of the processing data may include converting a meaningful display included in the information data into a corresponding meaning. The content may include at least one of text data, tables, graphs, images, audio information, or video information. The automatic content generation method may be stored in a computer-readable recording medium recording a program.

An apparatus for automatically generating content for achieving the technical problem includes an information storage unit including a cumulative data DB (Data Base) storing at least some of the past information data and a processed data DB storing the processed data, the information from an information provider system A pre-processing unit for receiving data, a processing data generation module for generating processing data based on at least one of the information related to the information data among the received information data or the information stored in the accumulated data DB, and the processing data generation module And a content generation unit generating content based on at least one of the processed data generated by the stored data, the information stored in the processed data DB, the information stored in the accumulated data DB, or the information data.

The preprocessing unit includes a preprocessing module that parses the received information data to extract at least a portion of the information data, and stores at least a portion of the extracted information data in the cumulative data DB, wherein the processed data generation module is extracted. The processed data may be generated based on at least one of the at least one piece of information data, data stored in the cumulative data DB, or at least one of real time data corresponding to the information data.

The preprocessing unit may further include a search word DB for storing a search word, and the preprocessing module may receive the information data corresponding to the search word from the information provider system based on the search word included in the search word DB.

In addition, the preprocessing module checks the validity of the received information data, and if valid, the at least one part of the information data may be extracted and stored in the cumulative data DB. The preprocessing module may extract at least a portion of the information data or check the validity of the content data based on predetermined form information.

The content generation unit includes a content generation module and a graphic data generation module, wherein the content generation module is generated as the content from at least one of information stored in the processed data DB, information stored in the cumulative data DB, or the information data. Determine content data that is a portion, the graphic data generation module converts the determined content data into at least one of a table or a graph, and the content generation module converts the determined content data into text data and converts the converted text Content including at least one of data, the table, and the graph may be generated.

The content generation module extracts a keyword from the determined content data, determines a term corresponding to the extracted keyword, determines numerical data when there is numerical data corresponding to the keyword, and reconstructs the term and the determined numerical data. Arrange and add sentence components based on the rearrangement results.

The content generation module extracts a keyword included in the content data, compares numerical data corresponding to the keyword, selects a term corresponding to a comparison result, and determines analysis information based on the selected term and the keyword. The sentence component may be determined to be included between each of the determined analysis information, the keyword, or the verb, and may be converted into text data by combining the determined sentence component, the analysis information, the keyword, and the verb.

The content generation unit may further include a translation module for converting at least a portion of the generated content into another language.

The method and apparatus for automatically generating content according to the present invention may receive information data, generate various additional information or analysis opinions related to the information data, and provide the information consumer with the information data, thereby providing useful high quality information. There is an effect that can be obtained.

In addition, since these various additional information or analysis opinions are automatically generated, it is possible to provide high quality information to information consumers as quickly as possible.

In addition, even when delivering the information data itself to the information consumer has an effect that can be provided using the interface, such as various tables or graphs that can increase the readability.

In addition, if there is a predetermined error in the information data can be filtered, there is an effect that can always provide reliable content.

In order to fully understand the present invention, the operational advantages of the present invention, and the objects achieved by the practice of the present invention, reference should be made to the accompanying drawings which illustrate preferred embodiments of the present invention and the contents described in the accompanying drawings.

Also, in this specification, when any one element 'transmits' data to another element, the element may transmit the data directly to the other element, or may be transmitted through at least one other element And may transmit the data to the other component.

Conversely, when one element 'directly transmits' data to another element, it means that the data is transmitted to the other element without passing through another element in the element.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. Like reference numerals in the drawings denote like elements.

2 is a schematic functional block diagram of an apparatus for automatically generating content according to an embodiment of the present invention.

Referring to FIG. 2, the apparatus 100 for automatically generating content according to an exemplary embodiment of the present invention may include a preprocessor 110, an information storage unit 120, a processed data generation module 130, and a content generation unit 140. Include.

Each component of the automatic content generating apparatus 100 is separately shown in the drawings to indicate that it may be functionally and logically separated, and means that it is physically necessarily a separate component or implemented as a separate code. no.

In the present specification, the module may mean a functional and structural combination of hardware for performing the technical idea of the present invention and software for driving the hardware. For example, the module may mean a logical unit of a predetermined code and a hardware resource for performing the predetermined code, and means a physically connected code or does not necessarily mean one kind of hardware. It can be easily inferred by the average expert in the art.

The preprocessing unit 110 may include a preprocessing module 111, and may further include a search word DB 112, form information 113, metadata DB 114, and / or a timer 115.

The information storage unit 120 may include a cumulative data DB 121 and / or processed data DB (123).

The content generation unit 140 may include a content generation module 141, and may further include a thesaurus 142 for generating a sentence, a graphic data generation module 143, and / or a translation module 144. .

In the present specification, the DB may mean a functional structural combination of software and hardware for storing information corresponding to each DB. The DB may be implemented as at least one table, and may further include a separate DBMS (Database Management System) for searching, storing, and managing information stored in the DB. In addition, it may be implemented in various ways such as a linked-list, a tree, a relational DB, and includes all data storage media and data structures capable of storing information corresponding to the DB.

The preprocessing module 111 included in the preprocessor 110 receives information data (for example, disclosure information) from the information provider system 200, and inspects the received information data to implement the technical idea of the present invention. Pre-processing, parsing the contents, or extracting the necessary parts. The information provider system 200 may be a stock exchange, a KOSDAQ, a financial supervisor, or the like, when the automatic content generation device 100 according to an embodiment of the present invention is applied to a news article generating device related to a stock. However, it can be easily inferred by the average expert in the technical field of the present invention that the content automatic generation apparatus 100 may be varied according to the field or system to which the automatic content generating device 100 is applied.

Then, the information data (eg, disclosure information) pre-processed by the preprocessing unit 110, that is, a necessary portion of the information data, may be stored in the cumulative data DB 121 included in the information storage unit 120. have.

The information stored in the cumulative data DB 121 includes all or part of information data of the past. Therefore, the actually required portion of the information data may mean new data that is not stored in the cumulative data DB 121. For example, as shown in FIG. 6, the information data (ie, the third quarter of 2006 business results disclosure of Handok Pharmaceuticals) also includes data included in the information information of the past (ie, second quarter results). And since the data (that is, the second quarter results) is already included in the cumulative data DB 121, the preprocessing unit 110 extracts only the necessary portion (that is, the third quarter results, 20) It may serve to store in the cumulative data DB (121).

When the pre-processing is performed so that at least a part of the information data is stored in the cumulative data DB 121, the cumulative data DB 121 includes historical information data related to the information data (3rd quarter of 2006, Dokdo Pharmaceutical Sales Performance). Business performance before the third quarter of 2006).

Therefore, the content automatic generating device 100 does not simply deliver or arrange and deliver the information data (Q3 2006 Q1 2006 sales results), but compares and / or analyzes the information with the past data. It can deliver useful information to consumers.

Such information useful to the information consumer may be generated only based on the current information data stored in the cumulative data DB 121 and the past information data itself, but the linkage between the information stored in the cumulative data DB 121 and / or Alternatively, the information may be newly generated through a predetermined processing action called an operation.

In this specification, new information generated by processing is defined as processing data. As a result, the information included in the cumulative data DB 121, that is, the cumulative data DB 121 is at least a systematic collection of information provided from the information provider system 200, while the processed data is not provided in the cumulative data. It can be used to include all kinds or forms of information. Such processing data 130 may be generated by the processing data generation module 130. In order to generate the processed data, the processed data generating module 130 may generate the data based on the information stored in the accumulated data DB 121, that is, the current information data and the past information data. In addition to the information contained within, any form of information may be utilized. For example, the processing data generation module 130 may generate processing data based on real-time data in addition to the accumulated data. For example, the real-time data may be current stock quote data when applied to stock-related content.

Such real-time data (eg, stock quote data) is also stored in a predetermined DB (not shown), so that not only current real-time data (eg, current stock quote data) but also historical real time data (eg, past stock quote data) Can be stored for inclusion. Accordingly, the processing data generation module 130 may use predetermined processing data (for example, price earnigs ratio (PER) or price book value ratio (PBR) based on current real-time data (stock price)). ), And based on past real-time data (historical stock quotes), how real-time data (stock quotes) can affect what is included in current information data (for example, operating results). It is also possible to generate processed data, which is information analyzed or processed based on (past stock quotes).

The processing data generation module 130 may store the generated processing data in the processing data DB 123. Therefore, not only the current processing data but also the past processing data may be stored in the processing data DB 123.

As a result, the information storage unit 120 may include not only the above-described cumulative or processed data set but also any kind or form of information that may be useful to the information consumer 300. For example, when the total number of images or voice information of a specific company is stored in the information storage unit 120, the image and / or voice information may be provided together when content related to the specific company is provided.

Meanwhile, the generated processing data may be useful information for the information consumer 300. The information consumer 300 may be, for example, a web client of a stock investor (eg, a home trading system (HTS) 310) or a news provider (eg, an internet newspaper, etc. 320). In addition, not only the processing data but also the comparison between the accumulated data stored in the cumulative data DB 121 (that is, the comparison between the current information data and the past information data) and / or the analysis information may also be useful to the information consumer 300. Can be. Alternatively, comparison and / or analysis information of past processed data and current processed data may also be useful information to the information consumer 300. Accordingly, all such information may be generated as content and provided to the information consumer 300.

The content may be used to include all kinds or forms of information having information meaningful to the information consumer 300. The content may include not only text but also images, tables, graphs, videos, and voices. It may be provided in the form. The type and form of the content may vary according to implementation. That is, the information that may be included in the content or its form may vary according to the type or form of data stored in the information storage unit 120.

As such, the content generating unit 140 may be a component that generates useful information generated based on various information such as the processed data, the accumulated data, and / or real time data in the form of content.

That is, the content generation unit 140 generates data that is fragmentary numerical values or symbols, such as some of the processing data generated by the processing data generation module 130 or the information stored in the information storage unit 120. As shown in FIG. 1B, a sentence may be generated and a table or a graph may be generated.

In addition, the content generation unit 140 may generate an analysis opinion based on a predetermined logic based on the information received from the information storage unit 120 and include it in the content.

The various useful information generated in this way, such as additional information or analysis opinions are automatically provided to the information consumer 300, thereby providing an information consumer with high quality information as quickly as possible.

Hereinafter, the function and role of each component will be described in more detail.

Referring back to FIG. 2, the preprocessing module 111 included in the preprocessor 110 performs various preprocessing functions as described above. That is, the preprocessing module 111 may parse at least the received information data to extract at least a portion of the information data and store at least a portion of the extracted information data in the cumulative data DB 121. have.

Meanwhile, the preprocessing module 111 may refer to a predetermined search word DB 112 to receive only necessary information from the information provider system 200 as the information data.

3 shows an example of a search word DB according to an embodiment of the present invention.

Referring to FIG. 3, in the search word DB 112, a predetermined search word may be stored in the search word DB 112 in order to receive desired information or necessary information by the content automatic generation device 100. Therefore, the preprocessing module 111 may receive only the information data corresponding to the search word from the information provider system 200 based on the search word included in the search word DB 112. This is because there are various types of information provided by the information provider system 200 and may include information that is not suitable for generating content directly. For example, the dynamic or promotional materials of a specific company may be generated and provided by the information provider system 200, and such information may be undesired information by the information consumer 300. Because it does not. As such, an example of various information generated and provided from the information provider system 200 is illustrated in FIG. 4. Referring to FIG. 4, information such as guidance of renaming and suspension may be information that is not suitable for generating content according to an embodiment. Of course, according to the embodiment of the automatic content generation device 100, it is obvious that the content can be generated using this information.

As a result, the pretreatment module 111 of the number 1040 Handok Pharmaceutical Co., Ltd. corresponding to a search term of a sales (provisional) record (process disclosure) included in the search word DB 112 of FIG. 3 among various pieces of information as shown in FIG. 4. Information data corresponding to business (provisional) performance (fair disclosure) will be received.

The search word stored in the search word DB 112 may vary according to the type of information provided by the information provider system 200 and / or an implementation example of the automatic content generating device 100.

Meanwhile, as illustrated in FIG. 3, a period in which the preprocessing module 111 searches for information data from the information provider system 200 may vary according to information data corresponding to each search word. For example, it is desirable to search in real time for events that may occur at any time, such as a single sale or supply contract (part number 01). On the other hand, the information data provided according to a certain period, such as the semi-annual report (section number 03) need only be periodically retrieved.

As such, the preprocessor 110 may further include a predetermined retrieval period control device, that is, a timer 115, to efficiently retrieve and receive the information data from the information provider system 200.

Meanwhile, the preprocessing module 111 may use a predetermined form to recognize the contents of the disclosure information provided in text or various forms, as shown in FIG. 1A. It may be stored in. For example, in the case of the performance disclosure as shown in FIG. 6, a form having only a field for some data 10 may be automatically input to the corresponding position in the form. Can be. In this way, the information entered in a specific form is easy to search, process, and calculation, and since only specific information comes to a specific location, it is possible to grasp the contents of the performance disclosure and easily extract only necessary parts. For example, when the specific form is written in Excel®, the partial data 10 may be input into each corresponding cell of Excel.

As described above, the content may be parsed, searched, processed, etc., but the preprocessing module 111 may also check the validity of the received information data.

For example, the inspection logic for performance disclosure among disclosure information includes longitudinal inspection and horizontal inspection of financial statements, and horizontal inspection is established by the manager on the premise that the company's sales are rarely changed suddenly QoQ. If it is out of range, it is recognized as an error. This error is often caused by mis-entering the unit when the company enters the disclosure (eg, when it is stated between 100 million won and 1 million won). The predetermined range of the above may be variously set according to the implementation example. Depending on the degree of such setting, it may be determined that valid information data is not valid for a company whose sales suddenly change abruptly. However, in such a case, other additional methods may be further used, such as supplementing the information data determined to be invalid by a verification procedure or the like by the manager or the management system. Longitudinal inspection is to check whether the value of the deduction between items using accounting principles is consistent. Of course, the validity of the inspection may vary depending on the kind of the actual information data.

As described above, only when it is determined that the information data or a part of the information data received from the information providing source system 200 is valid, the preprocessing module 111 transmits at least some of the information data to the cumulative data DB 121. Can be extracted and stored. Therefore, there is an effect that can increase the reliability of the content generated by the content automatically generating device 100.

The preprocessing unit 110 further includes a metadata DB 114 to store metadata such as a result of validity of the received information data or predetermined information (eg, whether or not processing data is generated) required in the preprocessing process. can do.

FIG. 5 shows an example of a metadata DB according to an embodiment of the present invention. As shown in FIG. 5, the preprocessing module 111 indicates a field indicating whether information data of the metadata DB 114 is valid. Whether the test result and / or the test result is valid is recorded, and according to the result, it is possible to determine whether to store the information data or at least part of the information data in the cumulative data DB 121. In addition, various tags (eg, whether processed data is generated, etc.) necessary in the preprocessing may be further included in the metadata DB 114.

6 is a diagram illustrating an example of a pretreatment process according to an exemplary embodiment of the present invention.

Referring to FIG. 6, the preprocessing unit 110 receives the information data (handok drug third quarter business (provisional) results (process disclosure)) received from the information provider system 200 and the specific portion 10 Figure out only. In addition, a portion (ie, third quarter performance, 20) to be additionally stored in the cumulative data DB 121 is extracted from the specific portion 10, and the extracted portion is accumulated data DB 121 having a predetermined format. ) Is stored in a specific table (30). Numerical data among the stored data 40 may be stored by changing its unit as necessary. In addition, in order to grasp only a specific portion 10 of the information data (Handok Pharmaceutical's third quarter sales (provisional) results (process disclosure)) or to store additional portions (ie, third quarter's performance, 20) As described above, the predetermined form information 113 can be used.

As such, the table 30 of the stored cumulative data DB 121 may include all past information data (past business results) related to the information data (current business results). Therefore, the cumulative data DB 121 may include a plurality of tables according to the type of received information data.

Then, the processing data generation module 130 may generate processing data based on the information stored in the accumulated data DB 121. Of course, other information (for example, real-time data) may be additionally required in addition to the information stored in the cumulative data DB 121.

7 is a diagram for describing generation of processing data according to an exemplary embodiment of the present invention.

Referring to FIG. 7, the processed data generation module 130 may generate processed data based on the information stored in the accumulated data DB 121. For example, the processed data generation module 130 may be interlocked by interlocking a table 121-1 having total capital data and a table 121-2 having net profit data among tables included in the cumulative data DB 121. Return on equity (ROE) can be calculated.

Alternatively, the PBR may be calculated based on the table 121-1 having the total capital data and real time data (eg, stock quote). Alternatively, the PER may be calculated based on the table 121-2 having the net profit data and real time data (eg, stock price).

The processed data may be meaningful data generated through a predetermined calculation or processing of information not included in the cumulative data DB 121, that is, information not provided from the information provider system 200.

The generated processing data may be stored in a predetermined table 123-1 of the processing data DB 123. The machining data may vary according to an embodiment, and may be stored in at least one table according to the type and characteristic of the machining data.

In addition, the processing data that can be generated by the processing data generation module 130 may vary depending on the type and form of the information stored in the information storage unit 120 and the additional information, such as the real-time data. As shown.

As shown in FIG. 7, the generated processing data is only numerical data corresponding to one keyword (eg, ROE, OP margin, PER, or PBR) or a keyword. However, the technical idea of the present invention generates and provides such keywords or numerical data as contents that can be easily utilized by the information consumer 300. For example, based on the information stored in the processed data table 123-1 and the information stored in the cumulative data DB 121, the apparatus 100 automatically generates the content 50 and / or the content (shown in FIG. 7). 60).

The content 50 includes text data on the ROE and OP margin among the generated processed data (ie, "Hando's annual ROE and OP margin including 1Q quarter results are 10.1% and 12.5%", etc.) and Corresponding graph is included.

The processed data used in the content 50 may be ROE, OP margin, peak of five quarters, and the like. As shown in FIG. Can be created as text data. In addition, a graph corresponding to the text data may be generated.

In addition, the content 60 includes text data and graphs based on the processed data PER and PBR. In particular, it can be seen that the content 60 also includes an analysis opinion based on the processed data ("In our case, the results compared to the same period last year."

As such, the content generation unit 140 may perform a function of generating the processed data into various types of contents such as not only text data and graphs but also tables and generating analysis opinions.

The content generation unit 140 includes a content generation module 141 and a graphic data generation module 143. In addition, the content generation unit 140 may further include a sentence generation thesaurus 142 and / or a translation module 144.

The content generation unit 140 may generate content based on the processed data as shown in FIG. 7, but based on the information stored in the accumulated data DB 121 and / or the received current information data itself. You can also create content. As such, a portion of the processed data to be stored in or stored in the processed data DB 123, information stored in the accumulated data DB 121, and / or information data to be generated as content may be defined as content data. What information may be determined as the content data may vary depending on the type and implementation of the received information data.

The content generation module 141 may determine content data and convert the determined content data into content. In particular, in the content generation unit 140 according to an embodiment of the present invention, the content generation module 141 is responsible for converting content data into text data or generating an analysis opinion corresponding thereto. The graphic data generating module 143 may separately perform a role of generating various graphic data such as an image and a graph. However, this may be various modifications if only one embodiment.

How the content generation module 141 converts the content data into text data may vary. One embodiment thereof is as follows.

For example, in the processed data table 123-1 shown in FIG. 7, it may be assumed that the item code (002390), the year (200609), the ROE (10.1), and the operating margin (12.5) are determined as content data. Then, the content generation module 141 extracts a keyword (item code, year, month, ROE, OP margin) from the content data, and corresponds to the term (Handok Pharm, Q3, ROE, OP margin) corresponding to the extracted keyword. ) Can be determined. In addition, when there is numerical data corresponding to the keyword, the numerical data (eg, ROE 10.1 and OP margin 12.5) may be determined. The term corresponding to the keyword may be determined based on the sentence generation thesaurus 142. An example of the sentence generation thesaurus 142 is illustrated in FIG. 9. The sentence generation thesaurus 142 may store a specific word or keyword, and various information such as a similar word, a sentence component, and a description thereof may be stored. . Using the thesaurus 142, it can be seen that sentences can be generated according to various embodiments.

Then, the term and the numerical data may be rearranged as necessary based on the term corresponding to the keyword and the determined numerical data. Of course, there may be cases where no rearrangement is necessary. Then, a predetermined additional word (eg, “one year standard”) may be determined. Such additional words may be preset according to the type of the information data. Then, text data can be generated by adding necessary sentence components (eg, search, ending, etc.) based on the terms, numerical data, and additional words. In order to add the sentence component, a suitable investigation or a mother may be selected using a state machine well known in the automata field of computational science.

The text data generated as described above may be included in the content 50 of FIG. 7, wherein the year-end ROE and operating margin are 10.1% and 12.5% of Handok Pharmaceuticals including 3Q results.

In addition, there may be an opinion of analysis among the text data generated based on the content data. The process of generating such an opinion is shown in FIGS. 8 and 9.

8 and 9 are diagrams for describing a process of converting content data into text data according to an embodiment of the present invention.

8 and 9, the content generation module 141 determines the sales and operating profit data 71 over two quarters of the predetermined table 70 included in the cumulative data DB 121 as the content data. Can be.

Then, the content generation module 141 may convert the content data into text data and generate the content as described above. However, the content generation module 141 may generate the predetermined analysis opinion as the content by analyzing the content data.

To this end, the content generation module 141 extracts a keyword included in the content data. For example, sales and operating profit may be keywords. Thereafter, the content generation module 141 may compare numerical data (561.7 billion and 638.8 billion) corresponding to the keyword (eg, sales) and select a word (eg, increase) corresponding to the comparison result. In addition, it is possible to compare numerical data (59.6 billion and 11.3.8 billion) corresponding to another keyword (eg, operating profit), and select a corresponding word (eg, increase) according to the comparison result. The content generation module may then determine analysis information (eg, the arrival of an active period) based on the selected verb (eg, all increase) and the keyword (eg, operating profit and sales).

Then, a sentence component to be included between each of the determined analysis information (eg, the arrival of the boom period), the keyword (eg, sales and operating profit), or the verb (increase, increase) is determined, and the determined sentence component and the analysis information are determined. , The keyword, and the term may be combined to generate text data corresponding to the opinion of the analysis (eg, “a sales are expected and operating profit is also expected to be a booming period”).

Meanwhile, in order to determine the gist, analysis information, and / or the sentence component, the content generation module 141 may use predetermined logic illustrated in FIG. 8.

For example, the content generation module 141 may first determine whether the sales increase by comparing the sales (S1), and then determine whether the operating profit has increased (S2). According to the second determination, a term C1 or C2, G1 or G2 corresponding to the keyword A1 or E1 determined and analysis information H1, H2, H3 or H4 may be determined. The sentence components b, d, and f may also be determined accordingly. The sentence component b1 corresponds to "this", d1 corresponds to "is", d2 corresponds to "but", and f1 and f2 correspond to "silver" and "degree", respectively.

Types of analysis comments that may be generated according to the sequential determination process are shown in reference numeral 80 of FIG. 9.

Meanwhile, in order to determine such keywords, terms, sentence components, and the like, the content generation module 141 may refer to the sentence generation thesaurus 142 as shown in FIG. 9. For example, when the word has the opposite meaning (eg, increase and decrease) in the above-described determination process by logic, it is possible to select an inverse ending (for example, but not referring to the note of the sentence generation thesaurus 142). will be.

One of the opinions generated in this way may be an opinion included in the content 60 of FIG. 7.

Meanwhile, the graphic data generating module 143 may generate the content data in a table or graph by performing a process of displaying numerical data included in the content data in a table or graph by a mathematical process. In addition, the user interface (UI) of the table or graph may vary depending on implementation.

10A to 10D are diagrams for describing information data and content data generated according to another embodiment of the present invention.

A description with reference to FIGS. 10A to 10D is as follows.

First, the automatic content generating apparatus 100 according to an embodiment of the present invention may receive disclosure information as shown in FIG. 10A from the information provider system 200.

Then, according to the conventional method, as shown in FIG. 10B, the AAA newspaper company created a news story 91 based on the disclosure information and provided it to the information consumer 300.

However, the apparatus 100 for automatically generating content according to an embodiment of the present invention receives the disclosure information illustrated in FIG. 10A, and automatically generates the contents as illustrated in FIGS. 10C to 10D to the information consumer 300. Can provide.

To this end, the preprocessor 110 may parse the disclosure information and extract a specific portion 90. The specific portion 90 may be stored in the cumulative data DB 121. Examples of the processed data and the contents generated based on the accumulated data DB 121 after storing and the information stored in the accumulated data DB 121 are illustrated in FIGS. 10C and 10D.

The preprocessing unit 110 parses the core content of the disclosure information, that is, disclosure data (eg, contracted goods, contracted amount, recent sales,% of sales, and counterparties), and the content generator 140 first The disclosure data included in the disclosure text may be converted into an article 93 in a sentence form. In addition, the processing data (eg, operating margin) and the like and / or the analysis opinion generated by the content generation unit 140 may be generated as text data 94 based on the disclosure data. .

In addition, the content generation unit 140 may generate the content 95 by generating the cumulative data for the corresponding item in the form of a table in the cumulative data DB 121. In addition, the content generation unit 140 may generate the content 96 by interlocking a search result for information having the same contract name in the cumulative data for another item, a stock price response (annual rate of change on the day of publication), and processing data (operating profit rate). Can be. In addition, the content generation unit 140 refers to a DB (not shown) for storing the cumulative data DB 121 and the real-time data for the search results and the stock price flow for the same information in the cumulative data for other items. You can also create

In addition, the content generation unit 140 may generate the content 98 by referring to a DB (not shown) that stores real time data separately from the information stored in the information storage unit 120.

As such, the automatic content generation device 100 according to an embodiment of the present invention generates a variety of useful content by using accumulated data, processed data, and additional information (for example, real-time data) to the information consumer 300. Can provide.

As a result, according to the conventional method, only the content corresponding to the information data is reported as it is, but the content automatic generation device 100 according to the embodiment of the present invention automatically generates the descriptive sentence form as well as the processed data or the analysis opinion. It is possible to provide multi-faceted information such as various application contents based on the cumulative data of the corresponding item and content using the cumulative data of other items in which the same event (for example, a contract) occurs.

In addition, there is an effect that can be extracted and displayed in the form of processed data of the third information that the information consumer did not think. By generating the content containing such processed data and providing it to the information consumer, the content consumer does not provide it in the information data, and thus the content that requires investment judgment to compete for minutes, such as various information (for example, investment information, etc.) that the information consumer must search and organize directly. The effect of the automatic content generating apparatus 100 may be maximized.

In addition, even if the information of the information data is not a number or a letter, but meaningful display, for example, '#', '*', '$', such as a symbol, symbol, symbol or special character pre-processing of the automatic content generating device 100 The preprocessing module 111 included in the unit 110 may perform a process of recognizing a symbol or a special character and converting the symbol or special character into a corresponding meaning. For example, when the symbol '+' is present, the term 'increase' may be converted into a term 'increase', or may be converted into a separate symbol used in the automatic content generating apparatus 100. For example, when a contract counterpart includes a string representing a stock company, such as a stock company or a 'Co.LTD', the content automatic generating device 100 may recognize the content by converting it into a meaning of 'stock company'. This may be possible by maintaining a separate synonym DB or by storing the synonym further in the sentence generation thesaurus 142.

FIG. 11 is a diagram for describing a process of converting at least a part of the generated content into a foreign language by a content generating device according to an embodiment of the present disclosure.

Referring to FIG. 11, the content generation unit 140 may further include a translation module 144 for converting at least some of the generated content into a foreign language, and the translation module 144 may include some of the generated content. (Eg, text data) can be converted into other languages.

To this end, the translation module 144 may receive the content generated from the content generation module 141 and translate it using a translator that is widely implemented. In addition, translation may be performed by further storing foreign language words corresponding to the words stored in the sentence generation thesaurus 142. In this way, the text data shown by reference numeral 80 of FIG. 9 can be translated into Japanese as shown in FIG.

The automatic content generation method according to an embodiment of the present invention can be embodied as computer readable codes on a computer readable recording medium. Computer-readable recording media include all kinds of recording devices that store data that can be read by a computer system. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, hard disk, floppy disk, optical data storage, and the like, and also in the form of carrier waves (e.g., transmission over the Internet). It also includes implementations. The computer readable recording medium can also be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion. And functional programs, codes and code segments for implementing the present invention can be easily inferred by programmers in the art to which the present invention belongs.

Although the present invention has been described with reference to one embodiment shown in the drawings, this is merely exemplary, and those skilled in the art will understand that various modifications and equivalent other embodiments are possible therefrom. Therefore, the true technical protection scope of the present invention will be defined by the technical spirit of the appended claims.

BRIEF DESCRIPTION OF THE DRAWINGS In order to better understand the drawings cited in the detailed description of the invention, a brief description of each drawing is provided.

FIG. 1A illustrates a conventional content generation method, and FIG. 1B illustrates an example of content generated according to an exemplary embodiment of the present invention.

2 is a schematic functional block diagram of an apparatus for automatically generating content according to an embodiment of the present invention.

3 shows an example of a search word DB according to an embodiment of the present invention.

4 shows an example of various information provided from an information provider system.

5 shows an example of a metadata DB according to an embodiment of the present invention.

6 is a diagram illustrating an example of a pretreatment process according to an exemplary embodiment of the present invention.

7 is a diagram for describing generation of processing data according to an exemplary embodiment of the present invention.

8 and 9 are diagrams for describing a process of converting content data into text data according to an embodiment of the present invention.

10A to 10D are diagrams for describing information data and content data generated according to another embodiment of the present invention.

FIG. 11 is a diagram for describing a process of converting at least a part of the generated content into a foreign language by a content generating device according to an embodiment of the present disclosure.

Claims (24)

Receiving, by the apparatus for automatically generating content, information data from the information provider system; Processing associated with the information data based on at least one of the received information data, a cumulative data database (DB) storing at least a portion of past information data related to the information data, or real-time data corresponding to the information data Generating data; And Generating content based on at least one of the generated processing data, past processing data, the cumulative data DB, or the information data, Receiving the information data, And receiving the information data corresponding to the search word from the information provider system based on a pre-stored search word. The method of claim 1, wherein generating the processing data comprises: Parsing the received information data to extract at least a portion of the information data, and storing at least a portion of the extracted information data in the accumulated data DB; And generating the processed data based on at least one of the extracted information data, data stored in the cumulative data DB, or at least one of the real-time data. delete Receiving, by the apparatus for automatically generating content, information data from the information provider system; Processing associated with the information data based on at least one of the received information data, a cumulative data database (DB) storing at least a portion of past information data related to the information data, or real-time data corresponding to the information data Generating data; And Generating content based on at least one of the generated processing data, past processing data, the cumulative data DB, or the information data, Generating the processing data, Validating the received information data; Parsing the received information data when valid, extracting at least some of the information data, and storing at least some of the extracted information data in the cumulative data DB; And generating the processed data based on at least one of the extracted information data, data stored in the cumulative data DB, or at least one of the real-time data. The apparatus of claim 4, wherein the apparatus for automatically generating content comprises: And extracting at least a part of the information data or validating the content data based on pre-stored form information. Receiving, by the apparatus for automatically generating content, information data from the information provider system; Processing associated with the information data based on at least one of the received information data, a cumulative data database (DB) storing at least a portion of past information data related to the information data, or real-time data corresponding to the information data Generating data; And Generating content based on at least one of the generated processing data, past processing data, the cumulative data DB, or the information data, Generating the content, Determining content data which is a part to be generated as the content from at least one of the generated processing data, the past processing data, the accumulated data DB, or the information data; And Converting the determined content data into at least one of text data, a table, and a graph, Converting the determined content data into at least one of text data, a table, or a graph, Extracting a keyword from the determined content data and determining a term corresponding to the extracted keyword; Determining numerical data when there is numerical data corresponding to the keyword; Rearranging the term, the numerical data determined; And Automatically generating content comprising the step of adding a sentence component based on the rearrangement result. delete Receiving, by the apparatus for automatically generating content, information data from the information provider system; Processing associated with the information data based on at least one of the received information data, a cumulative data database (DB) storing at least a portion of past information data related to the information data, or real-time data corresponding to the information data Generating data; And Generating content based on at least one of the generated processing data, past processing data, the cumulative data DB, or the information data, Generating the content, Determining content data which is a part to be generated as the content from at least one of the generated processing data, the past processing data, the accumulated data DB, or the information data; And Converting the determined content data into at least one of text data, a table, and a graph, Converting the determined content data into at least one of text data, a table, or a graph, Extracting keywords included in the content data and comparing numerical data corresponding to the keywords; Selecting a word corresponding to a comparison result and determining analysis information based on the selected word and the keyword; Determining a sentence component to be included between each of the determined analysis information, the keyword, or the verb, and converting the determined sentence component, the analysis information, the keyword, and the verb into text data. How to produce. The method of claim 1, wherein the automatic content generation method comprises: And converting at least a portion of the generated content into a foreign language. The method according to claim 1, 4, 6 or 8, wherein the processing data, A method of automatically generating content including at least one of a stock price return ratio, an operating profit margin, or a return on equity. The method according to claim 1, 4, 6 or 8, wherein the information data, How to automatically generate content that is a stock breaking news article. The method of claim 1, 4, 6 or 8, wherein the content is, A method for automatically generating content including at least one of text data, tables, graphs, images, audio information, or video information. The method of claim 1, 4, 6, or 8, wherein the generating of the processing data comprises: And converting the meaningful display included in the information data into a corresponding meaning. A computer-readable recording medium having recorded thereon a program for performing the method according to any one of claims 1, 2, 4, 5, 6, 8, or 9. An information storage unit including a cumulative data DB (Data Base) for storing at least a portion of past information data and a processed data DB for storing processed data; A preprocessor for receiving information data from the information provider system; Generate processing data for generating processing data related to the information data based on at least one of the received information data, information stored in the cumulative data DB, information related to the information data, or real time data corresponding to the information data module; And And a content generation unit generating content based on at least one of the processing data generated by the processing data generating module, information stored in the processing data DB, information stored in the cumulative data DB, or the information data. The preprocessing unit, A preprocessing module for parsing the received information data to extract at least a portion of the information data and storing at least a portion of the extracted information data in the cumulative data DB, The processing data generation module, And automatically generate the processed data based on at least one of the extracted information data, data stored in the cumulative data DB, or real time data corresponding to the information data. delete The method of claim 15, wherein the preprocessing unit, It further includes a search query DB that stores the search terms. And the preprocessing module is configured to receive the information data corresponding to the search word from the information provider system based on the search word included in the search word DB. The method of claim 17, wherein the pretreatment module, Validating the information data received, And automatically extracting and storing at least a portion of the information data in the cumulative data DB when the test result is valid. The method of claim 18, wherein the pretreatment module, And automatically extracting at least a portion of the information data or validating the content data based on predetermined form information. An information storage unit including a cumulative data DB (Data Base) for storing at least a portion of past information data and a processed data DB for storing processed data; A preprocessor for receiving information data from the information provider system; Generate processing data for generating processing data related to the information data based on at least one of the received information data, information stored in the cumulative data DB, information related to the information data, or real time data corresponding to the information data module; And And a content generation unit generating content based on at least one of the processing data generated by the processing data generating module, information stored in the processing data DB, information stored in the cumulative data DB, or the information data. The content generation unit, A content generation module; And Includes a graphic data generation module, The content generation module determines content data which is a part to be generated as the content from at least one of information stored in the processed data DB, information stored in the accumulated data DB, or the information data, The graphic data generation module converts the determined content data into at least one of a table or a graph, The content generation module converts the determined content data into text data and automatically generates content including at least one of the converted text data, the table, or the graph. The method of claim 20, wherein the content generation module, Extracting a keyword from the determined content data, determining a term corresponding to the extracted keyword, determining numerical data when there is numerical data corresponding to the keyword, rearranging the term, the determined numerical data, and rearranging. Automatic content generation device for adding a sentence component based on the result. The method of claim 20, wherein the content generation module, Extracts a keyword included in the content data, compares numerical data corresponding to the keyword, selects a word corresponding to a comparison result, determines analysis information based on the selected word and the keyword, and determines the determined analysis information And a sentence component to be inserted between each of the keyword and the verb and converting the determined sentence component, the analysis information, the keyword, and the verb into text data. An information storage unit including a cumulative data DB (Data Base) for storing at least a portion of past information data and a processed data DB for storing processed data; A preprocessor for receiving information data from the information provider system; Generate processing data for generating processing data related to the information data based on at least one of the received information data, information stored in the cumulative data DB, information related to the information data, or real time data corresponding to the information data module; And And a content generation unit generating content based on at least one of the processing data generated by the processing data generating module, information stored in the processing data DB, information stored in the cumulative data DB, or the information data. The content generation unit, And a translation module for converting at least a portion of the generated content into a foreign language. The method of claim 15, wherein the pretreatment module, Parsing the received information data and converting the meaningful display included in the information data into a corresponding meaning.

Images