Information processing device, information processing method and information processing program

Abstract

To address the requirement in a technical field of news clipping services, for specifying a place where an event occurs as a "place name" as well as providing "location information" of the place as additional information.SOLUTION: An information processing device includes an object document data acquisition unit (11) that acquires object document data to be analyzed, a place name information acquisition unit (12) that uses a large-scale language model (LLM) to acquire place name information of a place where an event represented by the object document data has occurred, a location information specification unit (13) that uses a place name location information database (PDB) to specify location information associated with the place name information, and a distribution information generating unit (15) that generates distribution content in which the object document data and the location information specified by the location information specification unit (13) are associated with each other.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to an information processing device, an information processing method, and an information processing program. [Background technology]

[0002] 2. Description of the Related Art For organizations or individuals engaged in various activities such as business activities or social activities, it is useful to know the occurrence of events that affect their own activities among various events occurring in the world. As a service that meets this need, so-called news clipping services are known, in which users set keywords of interest in advance and receive only news items that match the keywords.

[0003] There are also known technologies for providing customers with information including not only domestic news but also international news. For example, Patent Document 1, previously filed by the applicant of the present application, describes a technology for acquiring target document data to be analyzed, determining whether or not there is a risk due to an event indicated by the content of the target document data, and, if there is a risk, translating the data into a specific language to generate distribution information. Such a service can also be considered a global news clipping service. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent No. 7341367 Summary of the Invention [Problem to be solved by the invention]

[0005] According to the technology described in Patent Document 1, when generating distribution information written in natural language, it is possible to selectively extract place names contained in the distribution information and use the extracted place names as labels for the distribution information.

[0006] In the technical field of news clipping services, it is required not only to identify the place where an event occurred as a "place name" but also to provide the "location information" of that place as additional information. The disclosed technology aims to provide an information processing device, an information processing method, and an information processing program that improve on conventional news clipping services and not only identifies the location where an event occurred as a "place name," but also provides the "location information" of that location as additional information. [Means for solving the problem]

[0007] The information processing device according to the disclosed technology includes a target document data acquisition unit that acquires target document data to be analyzed, a place name information acquisition unit that uses a large-scale language model to acquire place name information of a location where an event indicated by the target document data occurred, a location information identification unit that uses a place name location information database to identify location information linked to the place name information, and a distribution information generation unit that generates distribution content that associates the target document data with the location information identified by the location information identification unit. [Effects of the Invention]

[0008] Because the information processing device according to the disclosed technology has the above configuration, it can not only identify the location where an event occurred as a "location name," but also provide the "location information" of that location as additional information. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 is a block diagram showing a functional configuration of an information processing device according to the first embodiment. [Figure 2] FIG. 2 is a block diagram showing a functional configuration of the information processing device according to the second embodiment. [Figure 3] FIG. 3 is a block diagram showing a functional configuration of an information processing device according to the third embodiment. [Figure 4] FIG. 4 is a block diagram showing the hardware configuration of an information processing device according to the disclosed technique. [Figure 5]FIG. 5 is a flowchart showing the processing flow of the information processing method according to the disclosed technique. DETAILED DESCRIPTION OF THE INVENTION

[0010] Embodiment 1 Fig. 1 is a block diagram showing the functional configuration of an information processing device 10 according to embodiment 1. As shown in Fig. 1, the information processing device 10 according to embodiment 1 includes a target document data acquisition unit 11, a place name information acquisition unit 12, a location information identification unit 13, and a distribution information generation unit 15. In Fig. 1, "LLM" represents the large-scale language model LLM, and "PDB" represents the place name location information database PDB.

[0011] <<Target document data acquisition unit 11>> The target document data acquisition unit 11 is a component for acquiring target document data to be analyzed. As described above, the information processing device 10 according to the disclosed technology can be said to realize a global news clipping service. Therefore, the target document data acquired by the target document data acquisition unit 11 is, for example, various news articles on the web. The news acquired by the target document data acquisition unit 11 may be only domestic news. In the following description, the target document data will be described assuming that it includes information on events overseas.

[0012] The target document data acquisition unit 11 may be configured to acquire target document data from specific information sources, such as reports published by public institutions, government information sites, overseas media, etc. The target document data acquisition unit 11 may also be configured to use external services, such as news distribution services, by using, for example, an API (Application Programming Interface).

[0013] The target document data acquisition unit 11 may be configured to scrape articles based on specific keywords. To clarify that scraping is performed from a website, "scraping" is often referred to as "Web scraping." For example, when disaster information is to be collected, the specific keywords used by the target document data acquisition unit 11 may be those related to disasters such as earthquakes, fires, typhoons, etc.

[0014] In addition, the target document data acquisition unit 11 may also have a function of acquiring information from social networking sites (SNS), such as X (formerly Twitter (registered trademark)), Instagram (registered trademark), TikTok (registered trademark), YouTube (registered trademark), or Facebook (registered trademark). In this case, the information processing device 10 according to the disclosed technology may use artificial intelligence to combat fake news and detect fake news.

[0015] <<Place name information acquisition unit 12>> The place name information acquisition unit 12 is a component that uses a large-scale language model (LLM) to acquire place name information related to an event indicated by target document data. The place name information acquired by the place name information acquisition unit 12 includes, for example, the place name of the place where the event occurred. Here, the place name includes the name of a country, the name of a city, the proper name of a facility, the general name of a facility, and the like. The place name information acquisition unit 12 inputs the target document data acquired by the target document data acquisition unit 11 into the large-scale language model LLM using a prepared prompt, and asks the large-scale language model LLM to extract and output place name candidates related to overseas event information. Here, the prompt is an instruction to cause the large-scale language model LLM to make predictions or suggestions. The prompt of the large-scale language model LLM is sometimes referred to as an AI prompt. The basic operation performed by the place name information acquisition unit 12 is to instruct the large-scale language model LLM to extract a place name. The prompt may also instruct the large-scale language model LLM to extract multiple place names. More simply, the place name information acquisition unit 12 causes the large-scale language model LLM to output place name words contained in news articles. Regardless of the language input to the large-scale language model LLM, the place name information acquisition unit 12 may instruct the large-scale language model LLM to output place name candidates in a specific predetermined language. In other words, the prompt used by the place name information acquisition unit 12 may specify the output language. The large-scale language model LLM may output multiple place names from one piece of target document data (or one news article obtained as target document data). That is, depending on the question input to the prompt of the large-scale language model LLM, multiple place names may be output from the prompt. The processing procedure in this case will become clear from the processing procedure of the re-determination processing unit 14 described below. The items extracted from the target document data include at least overseas event information and place names related to the extracted overseas event information. As described above, when extracting place names, an instruction may be given to estimate multiple place names. The input to the prompt may include, for example, an instruction to translate the extracted overseas event information and place names into a specific language. Furthermore, when extracting overseas event information, a category (event classification) may be specified before processing. Furthermore, the input to the prompt may include an instruction to check whether the extraction based on the event classification was appropriate.

[0016] At this stage, the information processing device 10 according to the disclosed technique links the overseas event information with candidate place names related to the overseas event information.

[0017] 《Position information identification unit 13》 The location information specifying unit 13 is a component that specifies location information linked to place name information using the place name location information database PDB. The place name location information database PDB used by the location information identification unit 13 has data in which place names are linked to location information for those place names. The place name location information database PDB used by the location information identification unit 13 may be a database held internally by the information processing device 10 according to the disclosed technology, or may be provided by an external service using an API. In other words, the location information identification unit 13 is a component that selects data related to overseas event information from the place name location information database PDB that contains data in which place names are linked to location information for those place names.

[0018] As described above, the place names assumed by the presently disclosed technology include country names, city names (prefecture names), municipal names, and proper names of facilities. Within the region of "country," there exists a region called "city," and within the region of "city," there exists a region called "municipalities" and a region called "facility." In this way, "country," "city," "municipalities," and "facility" can be said to have a hierarchical relationship. A "facility" is the name of a building or the like located at a specific latitude and longitude, such as an apartment building, park, or shopping mall. The name of a "facility" can be public or private. For example, examples of "country," "city," "municipalities," and "facility" include "Japan," "Tokyo," "Chiyoda Ward," and "Patent Office General Building," respectively. The "Patent Office General Building" is included in "Tokyo," which is included in "Japan." In this specification, a concept indicating a broader region is referred to as being at a higher level. In the above example, "Tokyo" is a place name that is hierarchically higher than "Japan Patent Office General Administration Building," and "Japan" is a place name that is hierarchically higher than "Tokyo." Note that "country," "city," "municipalities," and "facility" are merely examples. The information processing device 10 according to the disclosed technology may introduce new concepts of areas, such as "states" and "prefectures," to classify place names into a more multi-layered hierarchy. Furthermore, the information processing device 10 according to the disclosed technology may define the hierarchy of place names based on the address writing style adopted in the country where the overseas event information indicated by the overseas event information included in the target document data originates. Since the address of the Japan Patent Office General Administration Building is in Kasumigaseki, Chiyoda Ward, Tokyo, a hierarchical structure of "Tokyo," "Chiyoda Ward," and "Kasumigaseki" may be used. Therefore, it is preferable that the place name location information database PDB used by the location information identification unit 13 has the hierarchical structure information shown here. This is because the address systems used in each country allow locations to be identified based on hierarchical location concepts that are familiar to people in that area. For example, it is preferable that the place name location information database PDB used by the location information identification unit 13 also outputs hierarchical information such that when "Patent Office General Office" is entered as a search keyword, "Patent Office General Office" is included in "Tokyo Metropolitan Government."

[0019] The location information identified by the location information identification unit 13 using the place name location information database PDB is, for example, the latitude and longitude of a point that represents a place name. The point that represents a place name may be defined as appropriate, for example, the center of gravity of an area in which the place name is defined, or the entrance to a facility that represents the place name.

[0020] The location information identification unit 13 uses the place name location information database PDB to perform a search process using place name candidates obtained from the large-scale language model LLM as search keywords, and acquires location information associated with the place name. For example, if the keywords extracted by the large-scale language model LLM for a news item are "Japan Patent Office" and "Patent Office General Building," the location information identification unit 13 searches for "Japan Patent Office" and "Patent Office General Building" using the place name location information database PDB. The location information identification unit 13 can also use a "partial match" function in the search to extract "Japan" and "Japan" from "Japan Patent Office." The location information identification unit 13 can also use a concept search as a search algorithm to extract highly similar terms, such as the "National Center for Industrial Property Information and Training" (an organization related to the Japan Patent Office) in the above example. Note that partial match and concept search are just examples of search algorithms, and the search algorithms used in the disclosed technology are not limited to these. In general terms, the location information identification unit 13 searches the data stored in the place name location information database PDB for one or more search keywords, and extracts all hits regardless of the hierarchy of the location indicated by the hit location information. The information processing device 10 according to the present disclosure allows the service provider to pre-set the hierarchy of the place name associated with the overseas event information. In this case, the location information associated with the overseas event information is linked to the specified hierarchy of the place name. The lower the hierarchy of the place name (e.g., "facility"), the more detailed the information, but there is a risk that an incorrect place name unrelated to the news will be associated. For example, a hypothetical news item, or hypothetical facility name, such as "Kasumigaseki Hospital on Fire," may result in a hit for the facility name "Kasumigaseki Elementary School." In this case, the provider of the news clipping service according to the present disclosure may pre-set the hierarchy of the place name to the "city, ward, town, or village" level before actually providing the service. By setting the hierarchy of the place name to the "city, ward, town, or village" level in this way, the mistake of "Kasumigaseki Elementary School" being erroneously associated with the news item "Kasumigaseki Hospital on Fire" will be eliminated, and the correct place name at the "city, ward, town, or village" level, "Kasumigaseki," will be associated.

[0021] <<Distribution information generation unit 15>> The distribution information generating unit 15 is a component that generates distribution content in which the target document data is associated with the location information identified by the location information identifying unit 13 . The distribution content generated by the distribution information generation unit 15 is temporarily stored in a storage medium and distributed as appropriate in response to user requests. The distribution information generation unit 15 may manage the distribution content by linking, in addition to location information, the results of translation into a specified language, information on the date and time when an event occurred, and the like. A provider of a news clipping service according to the disclosed technology may not only manage the distribution content, but also actually distribute the linked information by including it in the distribution content. The distribution content is used for the news clipping service. The distribution content generated by the distribution information generation unit 15 may take various forms. For example, the distribution content may include only overseas event information extracted as all or part of the target document data. The distribution content may also include the event classification used to extract the overseas event information.

[0022] The distribution content is distributed by methods such as feed distribution, email distribution, or map display API. Feed distribution is an information distribution method that publishes a list of new information on a website, for example, in RSS (RDF Site Summary) data format. Map display API refers to using a map display service application using an API. For example, if the distribution content generated by the distribution information generation unit 15 uses the map display API and the news is related to the "Japan Patent Office General Headquarters," the distribution content will include an image of a map with an icon indicating the location of the Japan Patent Office General Headquarters. The information processing device 10 according to the disclosed technology is configured to allow each user to set keywords of interest in advance. The information processing device 10 according to the disclosed technology also allows each user to set regions of interest in advance. Keywords of interest and regions of interest can be set in combination. A condition combining keywords of interest and regions of interest is set for each user as a "delivery condition." For example, if a user is interested in semiconductor manufacturing, the delivery conditions may be set as "semiconductors," "Silicon Valley (Northern California, San Francisco Bay Area, USA)," or "TSMC." Furthermore, a company's product manufacturing base or local base may be registered as a use case for the delivery condition. Therefore, the regions that can be registered here are not limited to regions of interest, but may also be place names or locations related to the company's business, such as the user's company base. The information processing device 10 according to the disclosed technology may use information stored in advance in a storage medium, the information including information identifying each user, the distribution conditions of each user, and the distribution destination of each user. The information processing device 10 checks the distribution content against the distribution conditions of each user and distributes it to the destinations of users who satisfy the distribution conditions. Note that some users may not wish to receive the distribution content directly, but may instead simply receive a notification and download the content of their own volition as needed. In this case, the information processing device 10 may be configured to include a user notification unit 16, which will be described later in a third embodiment.

[0023] The information processing device 10 according to the present disclosure may set a region of interest by directly setting a location name or by setting a region within a predetermined distance from an arbitrary location. In this manner, a user can freely specify a location and a radial distance from the location. The information processing device 10 according to the present disclosure may be configured with a default value for the radial distance (e.g., 10 km). The user may access the information processing device 10 according to the present disclosure via a user terminal and register the region of interest. For example, the information processing device 10 may provide several options for the radial distance that the user can set (e.g., 1 km, 10 km, 100 km, etc.), allowing the user to select a desired value. The user may access the information processing device 10 according to the present disclosure via a user terminal and register distribution conditions, including the region of interest. The information processing device 10 checks the distribution conditions, including the area of ​​interest of each user, and distributes the distribution content to the distribution destinations of users who satisfy the distribution conditions. For example, if the distribution conditions are registered as the area of ​​interest of a certain user being "within a 10 km radius of point A," and information about an overseas event, such as a fire, is acquired within 10 km of point A, the distribution content is distributed to this user.

[0024] As described above, the information processing device 10 according to the first embodiment includes the components shown in FIG. 1, and therefore can not only identify the location where an event occurred as a "location name," but also provide the "location information" of that location as additional information.

[0025] Embodiment 2 The information processing device 10 according to the second embodiment is a modified example of the information processing device 10 according to the disclosed technique. Unless otherwise specified, the same reference numerals as those used in the first embodiment are used in the second embodiment. Furthermore, descriptions that overlap with those in the first embodiment are omitted as appropriate.

[0026] Fig. 2 is a block diagram showing a functional configuration of an information processing device 10 according to embodiment 2. As can be seen by comparison with Fig. 1 according to embodiment 1, the information processing device 10 according to embodiment 2 includes a redetermination processing unit 14 in addition to the components according to embodiment 1. As shown in FIG. 2, the redetermination processing unit 14 is disposed after the location information specifying unit 13 and before the distribution information generating unit 15.

[0027] <<Re-determination processing unit 14>> The redetermination processing unit 14 is a component that, when there are multiple pieces of data selected by the position information specifying unit 13, selects one piece of data using the large-scale language model LLM. As described above, the place name information acquisition unit 12 uses the large-scale language model LLM to acquire place name information related to an event indicated by the target document data. At this time, the large-scale language model LLM may intentionally or unintentionally extract multiple place names for one piece of target document data. Because each place has its own location, the extraction of multiple place names may result in the extraction of multiple location information (referred to as the "multiple location, multiple location case"). The large-scale language model LLM may also extract one place name for one piece of target document data. Even if there is only one place name, the location information identification unit 13 may extract multiple location information (referred to as the "single location, multiple location case"). In both the multiple location, multiple location case and the single location, multiple location case, the re-determination processing unit 14 performs re-determination processing. For example, when there are multiple place name candidates, the re-determination processing unit 14 may re-query the large-scale language model LLM to select one piece of data from the multiple place name candidates that is likely to represent the overseas event information contained in the target document data to be analyzed. The information input to the large-scale language model LLM prompt includes at least the original overseas event information or target document data, information on the multiple place name candidates, and question information inquiring about the correct place name related to the news among the multiple place name candidates. When the re-determination processing unit 14 re-queries the large-scale language model LLM, it obtains information from multiple levels of the place name location information database PDB and sends it as a set to the large-scale language model LLM. A specific example will be described below.

[0028] For example, suppose that when the place name information acquisition unit 12 inputs certain target document data into the large-scale language model LLM, the place name extracted by the large-scale language model LLM is "Chuo Ward." Then, suppose that when the location information identification unit 13 searches for "Chuo Ward" in the place name location information database PDB, four "Chuo Wards" are found: "A. Chuo Ward, Tokyo," "B. Chuo Ward, Sagamihara City," "C. Chuo Ward, Osaka City," and "D. Chuo Ward, Sapporo City." In such a case, the re-determination processing unit 14 re-queries the large-scale language model LLM to determine which of the four "Chuo Wards" (A to D) correctly represents the event in the target document data. The large-scale language model LLM can determine the appropriate location by analyzing the sentence, taking into account the context surrounding "Chuo Ward" that appears in the target document data, as well as other keywords, etc. It can be said that the four "Chuo Wards" A to D above can be distinguished by the place names of the next higher level to which each "Chuo Ward" belongs, namely, "Tokyo," "Sagamihara City," "Osaka City," and "Sapporo City." Location information can also be used as a means of distinguishing the four "Chuo Wards" A to D above. When multiple "Chuo Wards" are extracted from the candidate name "Chuo Ward," the re-determination processing unit 14 adds new information of other hierarchical levels (for example, "Tokyo") to the information of the candidate name initially extracted by the large-scale language model LLM, such as "A. Chuo Ward, Tokyo," "B. Chuo Ward, Sagamihara City," "C. Chuo Ward, Osaka City," and "D. Chuo Ward, Sapporo City," and inputs this information into the large-scale language model LLM for re-determination. The technology disclosed herein is equipped with this re-determination processing unit 14, and therefore the second determination based on the large-scale language model LLM is performed correctly.

[0029] The distribution information generating unit 15 according to the second embodiment generates distribution content in which the target document data is associated with the location information that has been re-determined by the re-determination processing unit 14 to be an appropriate location. As described above, the distribution content generated by the distribution information generation unit 15 is temporarily stored in a storage medium and distributed as appropriate in response to a user request. Note that the distribution information generation unit 15 may manage the distribution content by linking the result of translation into a predetermined language, information on the date and time when an event occurred, and the like, in addition to the location information.

[0030] As described above, the information processing device 10 according to the second embodiment includes the components shown in FIG. 2, and therefore can not only identify the location where an event occurred as a "location name," but also provide, as additional information, "location information" that has been re-determined by the large-scale language model LLM to be appropriate for the location of that location.

[0031] Embodiment 3 The information processing device 10 according to the third embodiment is a modified example of the information processing device 10 according to the disclosed technique. Unless otherwise specified, the same symbols as those used in the previous embodiments are used in the third embodiment. Furthermore, descriptions that overlap with those in the previous embodiments are omitted as appropriate.

[0032] Fig. 3 is a block diagram showing a functional configuration of an information processing device 10 according to embodiment 3. As can be seen by comparison with Fig. 1 according to embodiment 1, the information processing device 10 according to embodiment 3 includes a user notification unit 16 in addition to the components according to embodiment 1. As shown in FIG. 3, the user notification unit 16 is disposed after the distribution information generation unit 15.

[0033] <<User notification section 16>> The user notification unit 16 is a component that determines whether the location information identified by the location information identification unit 13 satisfies each of the notification conditions (hereinafter referred to as "notification conditions") specified for each user, and if it is determined that the location information satisfies the notification conditions, notifies the notification destinations linked to the notification conditions. Note that the term "distribution" means transmitting the distribution content generated by the distribution information generation unit 15, and the term "notification" means notifying with a small amount of information that there is content that can be distributed. The information processing device 10 according to the disclosed technology may be configured to allow each user to set a region of interest. With such a configuration, the user can be notified promptly of news related to the set region of interest. Possible methods for specifying a region of interest that can be set include directly specifying the name of a place and specifying an area within a predetermined distance from an arbitrary point.

[0034] The "notification conditions" used by the user notification unit 16 may be completely the same as the above-mentioned "delivery conditions", or may be partially the same. The user notification unit 16 notifies the notification destination before the content is distributed. Some users may not want the content to be distributed directly, but may want to receive a notification first and then download the content as needed at their own discretion. Users with such a desire should set "notification conditions." Conversely, users who want the content to be distributed as soon as possible should set "distribution conditions."

[0035] As described above, the information processing device 10 according to the third embodiment includes the components shown in FIG. 3, and therefore, in addition to the effects shown in the first embodiment, it has the effect of enabling the user to be notified promptly of news related to the area of ​​interest that the user has set.

[0036] The information processing device according to the disclosed technique may be realized as an information processing program that causes a computer to execute an information processing method.

[0037] 4 is a block diagram showing the hardware configuration of an information processing device 10 according to the present disclosure. As shown in FIG. 4, the hardware of the information processing device 10 includes a communication interface 100, an input / output interface 101, a processor 102, and a memory 103.

[0038] 5 is a flowchart showing the processing flow of the information processing method according to the disclosed technology. As shown in FIG. 5, the processing flow of the information processing method according to the disclosed technology includes ST11 performed by the target document data acquisition unit 11, ST12 performed by the place name information acquisition unit 12, ST13 performed by the location information identification unit 13, ST14a and ST14b performed by the redetermination processing unit 14, ST15 performed by the distribution information generation unit 15, and ST16a and ST16b performed by the user notification unit 16.

[0039] The functions of the target document data acquisition unit 11, place name information acquisition unit 12, location information identification unit 13, redetermination processing unit 14, distribution information generation unit 15, and user notification unit 16 in the information processing device 10 are realized by processing circuits. That is, the information processing device 10 includes processing circuits for implementing ST11 of the target document data acquisition unit 11, ST12 of the place name information acquisition unit 12, ST13 of the location information identification unit 13, ST14a and ST14b of the redetermination processing unit 14, ST15 of the distribution information generation unit 15, and ST16a and ST16b of the user notification unit 16. The processing circuit is a processor 102 (also referred to as a CPU, central processing unit, processing unit, arithmetic unit, microprocessor, microcomputer, or DSP) that executes an information processing program stored in a memory 103.

[0040] The functions of the target document data acquisition unit 11, the place name information acquisition unit 12, the location information identification unit 13, the re-determination processing unit 14, the distribution information generation unit 15, and the user notification unit 16 are realized by software, firmware, or a combination of software and firmware. The software and firmware are written as programs and stored in the memory 103. The processor 102 realizes the functions of each unit by reading and executing the programs stored in the memory 103. That is, the information processing device 10 includes the memory 103 for storing programs that, when executed by the processor 102, result in the execution of ST11 of the target document data acquisition unit 11, ST12 of the place name information acquisition unit 12, ST13 of the location information identification unit 13, ST14a and ST14b of the re-determination processing unit 14, ST15 of the distribution information generation unit 15, and ST16a and ST16b of the user notification unit 16. It can also be said that these programs cause a computer to execute the procedures or methods of the target document data acquisition unit 11, the place name information acquisition unit 12, the location information identification unit 13, the re-determination processing unit 14, the distribution information generation unit 15, and the user notification unit 16. Here, the memory 103 may be, for example, a non-volatile or volatile semiconductor memory such as RAM, ROM, flash memory, or EPROM. The memory 103 may also be in a form including a disk such as a magnetic disk, a flexible disk, an optical disk, a compact disk, a minidisk, or a DVD. Furthermore, the memory 103 may be in a form of an HDD or an SSD. As shown in FIG. 5, the processing steps realized by the information processing program according to the disclosed technology start with the target document data acquisition unit 11 acquiring target document data (ST11). Then, the location name information acquisition unit 12 acquires location name information (ST12). Then, the location information identification unit 13 identifies location information (ST13). Then, the redetermination processing unit 14 determines whether multiple pieces of identified location information exist (ST14a). If multiple pieces of identified location information exist, the redetermination processing unit 14 performs redetermination processing (ST14b). Then, the distribution information generation unit 15 generates distribution information (distribution content) (ST15). Then, the user notification unit 16 determines which users should be notified of the generated distribution information (distribution content) and whether the notification conditions for each user are met (ST16a). The user notification unit 16 notifies users who meet the notification conditions that distribution information (distribution content) exists.

[0041] As described above, the information processing method and information processing program according to the disclosed technology can realize a global news clipping service that not only identifies the location where an event occurred as a "place name," but also provides the "location information" of that place as additional information.

[0042] The information processing device according to the disclosed technology includes a target document data acquisition unit that acquires target document data to be analyzed, a place name information acquisition unit that uses a large-scale language model to acquire place name information of a location where an event indicated by the target document data occurred, a location information identification unit that uses a place name location information database to identify location information linked to the place name information, and a distribution information generation unit that generates distribution content that associates the target document data with the location information identified by the location information identification unit. With this configuration, the information processing device according to the disclosed technology can not only identify the location where an event occurred as a "location name," but also provide the "location information" of that location as additional information.

[0043] In one aspect of the information processing device according to the disclosed technology, the target document data acquisition unit acquires target document data regardless of the type of language in which the target document data is written, the place name information acquisition unit uses a large-scale language model that handles multiple types of languages, and the distribution information generation unit generates distribution content in a pre-specified type of language. Because of these technical features, the information processing device according to the disclosed technology can handle news in multiple languages.

[0044] In another aspect of the information processing device according to the disclosed technique, data may be hierarchically organized in a place name location information database used by the location information identification unit. With such technical features, the information processing device according to the disclosed technique can provide location information for place names at an appropriate level.

[0045] Another aspect of the information processing device according to the present disclosure is that the location information identification unit may select data from a pre-specified hierarchy from the data obtained from the place name location information database, and identify location information linked to the place name information. Because of these technical features, the information processing device according to the disclosed technique can provide location information for place names at a pre-specified level.

[0046] Another aspect of the information processing device according to the disclosed technique may include a redetermination processing unit that selects one piece of data using a large-scale language model when there are multiple pieces of data selected by the position information identification unit. Due to these technical features, the information processing device according to the disclosed technology can provide reliable location information related to news even when there are multiple location information candidates estimated to be related to news.

[0047] Another aspect of the information processing device according to the disclosed technology may include a user notification unit that determines whether the location information identified by the location information identification unit satisfies each of the conditions set for each user, and sends a notification to a notification destination associated with the condition if it is determined that the location information satisfies the condition. Because of these technical features, the information processing device according to the disclosed technique can notify each user in advance that there is news content to be distributed that may interest that user.

[0048] Another aspect of the information processing device according to the disclosed technology may include a method in which the user notification unit specifies an area of ​​interest that can be set for each user, by directly specifying the area by place name, and by specifying the area as an area within a predetermined distance from any point. Because of these technical features, a user of an information processing device according to the disclosed technology can freely specify a point of interest and a radial distance centered on that point.

[0049] The information processing method according to the presently disclosed technology may include a target document data acquisition unit acquiring target document data to be analyzed, a place name information acquisition unit acquiring place name information of a location where an event indicated by the target document data occurred using a large-scale language model, a location information identification unit identifying location information linked to the place name information using a place name location information database, and a distribution information generation unit generating distribution content that associates the target document data with the location information identified by the location information identification unit. Because it includes such method steps, the information processing method according to the disclosed technology can not only identify the location where an event occurred as a "location name," but also provide the "location information" of that location as additional information.

[0050] The information processing program according to the disclosed technique may be a program that causes a computer to execute the above-described information processing method. To execute the above procedure, the information processing program according to the disclosed technology can not only identify the location where the event occurred as a "location name," but also provide the "location information" of that location as additional information. [Industrial Applicability]

[0051] The disclosed technology can be applied to, for example, a global news clipping service, and has industrial applicability. [Explanation of symbols]

[0052] 10 Information processing device, 11 Target document data acquisition unit, 12 Place name information acquisition unit, 13 Location information identification unit, 14 Redetermination processing unit, 15 Distribution information generation unit, 16 User notification unit, 100 Communication interface, 101 Input / output interface, 102 Processor, 103 Memory, LLM Large-scale language model, PDB Place name location information database.

Claims

1. a target document data acquisition unit for acquiring target document data to be analyzed; a place name information acquisition unit that acquires place name information of a place where an event indicated by the target document data occurred using a large-scale language model; a location information specifying unit that specifies location information associated with the place name information using a place name location information database; a distribution information generation unit that generates distribution content in which target document data is associated with the location information identified by the location information identification unit, Information processing device.

2. the target document data acquisition unit acquires the target document data regardless of the type of language in which the target document data is written, the place name information acquisition unit uses the large-scale language model that handles a plurality of types of languages, the distribution information generation unit generates the distribution content in a predetermined language type; The information processing device according to claim 1 .

3. The place name location information database used by the location information identification unit is managed in a hierarchical manner. The information processing device according to claim 1 .

4. the location information specifying unit selects data of a predetermined hierarchy from the data obtained from the place name location information database, and specifies location information linked to place name information; The information processing device according to claim 3 .

5. a re-determination processing unit that selects one piece of data by using the large-scale language model when there are multiple pieces of data selected by the location information identification unit; The information processing device according to claim 1 .

6. and a user notification unit that determines whether the location information identified by the location information identification unit satisfies each of the conditions set for each user, and when it is determined that the location information satisfies the conditions, sends a notification to a notification destination associated with the conditions. The information processing device according to claim 1 .

7. The user notification unit specifies the area of ​​interest that can be set for each user, including a method of directly specifying the area of ​​interest by a place name and a method of specifying the area of ​​interest as an area within a predetermined distance from an arbitrary point. The information processing device according to claim 6 .

8. An information processing method for an information processing device including a target document data acquisition unit, a place name information acquisition unit, a location information identification unit, and a distribution information generation unit, the target document data acquisition unit acquires target document data to be analyzed, the place name information acquisition unit acquires place name information of a place where an event indicated by the target document data occurred using a large-scale language model; the location information specifying unit specifies location information linked to the place name information using a place name location information database; the distribution information generation unit generates distribution content in which the target document data is associated with the location information identified by the location information identification unit; Information processing methods.

9. An information processing program for causing a computer to execute the method according to claim 8.

Images