JP7639730B2 - How to search - Google Patents

Description

This disclosure relates to a search method.

Patent Document 1 proposes a word processing device for searching various materials. Specifically, the word processing device proposed in Patent Document 1 includes a search target document designation means, a word extraction means, an unprocessed word selection means, a word creation means, and a search means. The search target document designation means designates a document to be searched. The word extraction means extracts each word from the document designated as the search target using a predetermined method. The unprocessed word selection means selects a word of a predetermined length or more from among unprocessed words that could not be extracted by the word extraction means. The word creation means creates multiple words based on the selected unprocessed words, each with one character deleted from the beginning. The search means uses the created words as words for a forward match search.

JP 2002-351918 A

The purpose of this disclosure is to provide technology that improves the searchability of text.

A search method according to one aspect of the present disclosure is a search method executed by a computer, and includes acquiring a search term, acquiring structural information indicating a hierarchical relationship among a plurality of hierarchical sentences including a search target sentence, identifying a higher-level sentence that is located in a higher level than the search target sentence from among the plurality of sentences by referring to the acquired structural information, acquiring tag information assigned to the identified higher-level sentence, searching for a word that matches the acquired search term in the tag information of the search target sentence and the higher-level sentence, and outputting information related to the search result.

This disclosure can improve the searchability of text.

FIG. 1 illustrates an example of a situation to which the present disclosure is applied. FIG. 2 is a schematic diagram illustrating an example of a hardware configuration of an information processing device according to an embodiment. FIG. 3 is a diagram illustrating an example of a software configuration of the information processing device according to the embodiment. FIG. 4 is a flowchart illustrating an example of a processing procedure related to a text search performed by the information processing device according to the embodiment. FIG. 5 is a schematic diagram showing an example of a meta-information database according to an embodiment. FIG. 6 is a schematic diagram showing an example of configuration information and hierarchical text according to an embodiment. FIG. 7 shows an example of a comparison range after a process of acquiring tag information of a higher-level sentence for each sentence shown in FIG. 6 is executed.

In conventional search methods, whether or not to extract the search target sentence as a search result is determined depending on whether or not a word matching a specified search term is present in the sentence to be searched (search target sentence). In other words, if the search target sentence does not contain a word matching a search term, the search target sentence will not be extracted as a search result. Therefore, in order to extract a desired sentence, search terms are specified by predicting the words contained in the desired sentence. Therefore, conventional methods have the problem that the retrievability of sentences is not high (i.e., it is difficult to accurately extract a desired sentence).

In contrast, the search method according to the first aspect of the present disclosure is a search method executed by a computer, and includes obtaining a search term, obtaining structural information indicating a hierarchical relationship among a plurality of hierarchical sentences including the search target sentence, identifying a higher-level sentence that is located in a higher level than the search target sentence from among the plurality of sentences by referring to the obtained structural information, obtaining tag information assigned to the identified higher-level sentence, searching for a word that matches the obtained search term in the tag information of the search target sentence and the higher-level sentence, and outputting information related to the search result.

In the first aspect of the present disclosure, whether or not to extract the search target sentence as a search result is determined depending on whether or not there is a word matching the search term in the tag information of the search target sentence and the upper hierarchical sentence. In other words, the range for searching whether or not there is a word matching the search term (search term matching range) is expanded not only to the search target sentence but also to the tag information attached to the upper hierarchical sentence located in the upper hierarchical level of the search target sentence. As a result, even if the search target sentence does not contain a word matching the search term, if the tag information of the upper hierarchical sentence contains a word matching the search term, the search target sentence can be extracted as a search result. Therefore, according to the first aspect of the present disclosure, it is possible to increase the possibility that the target sentence meets the search criteria (i.e., it is possible to increase reproducibility). As a result, it is possible to increase the searchability of the sentence.

The search method according to the second aspect of the present disclosure may be the search method according to the first aspect, further including the computer acquiring tag information assigned to the search target sentence. The searching may be performed by searching the search target sentence, the tag information of the search target sentence, and the tag information of the higher-level sentence to determine whether or not there is a word matching the search term. With this configuration, the matching range of the search term can be further expanded to the tag information of the sentence itself. This can further improve the searchability of the sentence.

The search method according to the third aspect of the present disclosure may further include the computer acquiring one or more related words related to the search term. The searching may be configured by searching for the presence or absence of a word that matches at least one of the acquired search term or the acquired one or more related words. In this configuration, the words to be matched with the words contained in the search target sentence and tag information are expanded to not only the search term but also related words related to the search term. In other words, the search conditions can be relaxed so that related words are used as additional search terms. As a result, even if the search target sentence does not contain a word that matches the search term, if it contains a word that matches the related word, it is possible to extract the search target sentence as a search result. Therefore, according to this configuration, it is possible to increase the possibility of extracting the desired sentence without comprehensively specifying search terms. This saves the effort required to specify search terms and further improves the searchability of sentences.

A search method according to a fourth aspect of the present disclosure is the search method according to the third aspect, wherein the one or more related words may include at least one of a synonym, a similar word, a higher conceptual term, and a lower conceptual term of the search word. With this configuration, related words related to the search word can be appropriately acquired. As a result, an improvement in the search accuracy (match rate) of a sentence can be expected.

A search method according to a fifth aspect of the present disclosure is the search method according to the third or fourth aspect, and may be configured such that acquiring the one or more related words includes acquiring a category of the search term and acquiring one or more related words belonging to the acquired category. According to this configuration, since the related words are categorized, it is possible to make it difficult to acquire related words that are less relevant to the search term in relation to the text to be searched. In other words, in a search scene, it is possible to appropriately acquire related words that are highly relevant to the search term. As a result, it is possible to expect an improvement in the search accuracy (match rate) of texts.

An embodiment according to one aspect of the present disclosure (hereinafter also referred to as "the present embodiment") will be described below with reference to the drawings. However, the present embodiment described below is merely an example of the present disclosure in all respects. Various improvements or modifications may be made without departing from the scope of the present disclosure. In implementing the present disclosure, a specific configuration according to the embodiment may be appropriately adopted. Note that while the data appearing in this embodiment is described in natural language, more specifically, it is specified in pseudo-language, commands, parameters, machine language, etc. that can be recognized by a computer.

[1 Application Examples]
1 is a schematic diagram illustrating an example of a situation in which the present disclosure is applied. An information processing device 1 according to the present embodiment is one or more computers configured to execute a search process for extracting sentences that meet search criteria.

In this embodiment, the information processing device 1 acquires a search term 5. The information processing device 1 also acquires structural information 200 indicating a hierarchical relationship among a plurality of hierarchically organized sentences 2 including a search target sentence 20. The information processing device 1 identifies, from among the plurality of sentences 2, a higher-level sentence 25 that is located in a higher level than the search target sentence 20, by referring to the acquired structural information 200. The information processing device 1 acquires tag information 35 assigned to the identified higher-level sentence 25. The information processing device 1 searches for a word matching the acquired search term 5 in the tag information 35 of the search target sentence 20 and the higher-level sentence 25. The information processing device 1 then outputs information related to the search results.

As described above, in this embodiment, the matching range of the search term 5 is expanded not only to the search target sentence 20 but also to the tag information 35 attached to the upper level sentence 25 located in a higher level than the search target sentence 20. That is, in this embodiment, the matching range is composed of the search target sentence 20 and the tag information 35. As a result, even if the search target sentence 20 does not contain a word matching the search term 5, if the tag information 35 of the upper level sentence 25 contains a word matching the search term 5, the search target sentence 20 can be extracted as a search result. Therefore, according to this embodiment, it is possible to increase the possibility that the target sentence satisfies the search criteria.

In addition, tags associated with a sentence generally often succinctly indicate information such as the attributes and characteristics of the sentence. Furthermore, in sentences having a hierarchical structure, sentences located at higher levels of the hierarchy often indicate the main points and summaries of sentences located at lower levels. Therefore, the tag information 35 of the higher-level sentence 25 is likely to include, as tags, words that are highly related to the search target sentence 20 located at lower levels in the hierarchical structure and that are likely to be used as search conditions. By adding this tag information 35 to the matching range instead of the higher-level sentence 25 itself, it is possible to make it easier to appropriately extract the search target sentence 20 while suppressing the amount of expansion of the matching range. Therefore, according to this embodiment, it is possible to efficiently improve the searchability of sentences.

(multiple sentences)
In addition, as long as the multiple documents 2 have a hierarchical structure, the type of the multiple documents 2 need not be particularly limited and may be appropriately selected depending on the embodiment. The multiple documents 2 may be, for example, legal documents, specifications, etc. The multiple documents 2 may be obtained from at least a portion of one or more documents. In one example, the documents may correspond to units extracted by search. The hierarchical structure may be given, for example, by chapters, bullet points (articles, paragraphs, numbers, etc.), etc. As long as the structure information 200 is configured to indicate the hierarchical structure in the multiple documents 2, the data format thereof need not be particularly limited and may be appropriately selected depending on the embodiment.

(Text to search)
The search target sentence 20 is a sentence that is the subject of the search process. The search target sentence 20 may be arbitrarily selected from a plurality of sentences 2. In one example, the search target sentence 20 may be specified by a user. A plurality of search target sentences 20 may be acquired by specifying a document, a range, etc. to be searched. When a plurality of search target sentences 20 are acquired, the information processing device 1 may execute a process of searching for each of the search target sentences 20 to determine whether or not a word matching the search term 5 exists. In another example, one or more search target sentences 20 may be determined in advance within the plurality of sentences 2. Each of all sentences 2 may be treated as a search target sentence 20.

(Matching range)
In this embodiment, the information processing device 1 may further expand the matching range to the tag information 30 assigned to the search target sentence 20 itself. That is, the information processing device 1 may further acquire the tag information 30 assigned to the search target sentence 20. The matching range may be composed of the search target sentence 20, the tag information 30 of the search target sentence 20, and the tag information 35 of the upper hierarchical sentence 25. In the search process, the information processing device 1 may search for a word matching the search term 5 in the search target sentence 20, the tag information 30 of the search target sentence 20, and the tag information 35 of the upper hierarchical sentence 25. This can further improve the searchability of the sentence.

The number of upper-level sentences 25 may be any number. One or more upper-level sentences 25 may exist in the upper level of the sentence 20 to be searched. When multiple upper-level sentences 25 exist, in one example, the information processing device 1 may be configured to acquire tag information 35 of all the upper-level sentences 25. In another example, the information processing device 1 may be configured to acquire tag information 35 of a portion of the multiple upper-level sentences 25 (e.g., a predetermined number in order from the closest hierarchical level).

(Tag information)
The tag information (30, 35) is composed of one or more tags associated with the sentence to which it is assigned (search target sentence 20, upper-level sentence 25). A tag is a word that indicates information such as attributes and characteristics of the associated sentence. A word (e.g., a word, a character string, etc.) contained in the associated sentence may be used as a tag. In one example, the tag may be generated manually. In another example, the tag may be generated by computer information processing.

(Search process)
In the search process, the information processing device 1 determines whether or not to extract the search target sentence 20 depending on whether or not a word matching the search term 5 is present in the collation range. If a word matching the search term 5 is present in the collation range, the information processing device 1 extracts the search target sentence 20 as a search result (i.e., a sentence that meets the search conditions). On the other hand, if a word matching the search term 5 is not present in the collation range, the information processing device 1 does not extract the search target sentence 20.

(Search term)
Typically, the search term 5 may be specified by a user input. However, the method of specifying the search term 5 is not limited to this example, and may be determined appropriately depending on the embodiment. In another example, an arbitrary word may be selected in an arbitrary sentence, and the selected word may be used as the search term 5. In other words, the search term 5 may be specified by selecting a word in an arbitrary sentence.

The number of search terms 5 may be any number. Multiple search terms 5 may be obtained. In this case, each search term 5 may be specified as a condition for a logical sum (OR) search, or may be specified as a condition for a logical product (AND) search. Some of the multiple search terms 5 may be specified as conditions for a logical sum search, and the rest may be specified as conditions for a logical product search.

(Expanding search criteria)
In this embodiment, the information processing device 1 may be configured to use a related word 6 related to the search term 5 as an additional search term. That is, the information processing device 1 may further acquire one or more related words 6 related to the search term 5. In the search process, the information processing device 1 may search for whether or not a word matching at least one of the acquired search term 5 or one or more related words 6 exists in the matching range.

In this way, by using related words 6 as additional search terms, the search conditions can be expanded. As a result, even if the search target sentence 20 does not contain a word that matches search term 5, it can be extracted as a search result if it contains a word that matches related words 6. Therefore, it is possible to easily extract a desired sentence without comprehensively specifying search terms 5. This saves the effort of specifying search terms 5 and further improves the searchability of sentences.

In one example, the related words 6 may be used as a condition for a logical sum search. However, the manner in which the related words 6 are used need not be limited to this example. In another example, when multiple related words 6 exist, at least some of the multiple related words 6 may be used as a condition for a logical product search. When multiple search terms 5 are specified, the information processing device 1 may obtain one or more related words 6 for at least some of the multiple search terms 5.

In this embodiment, the related words for each word may be stored in a meta-information database 60. As long as the meta-information database 60 is configured to be able to acquire related words for a specified word (target word), its format need not be particularly limited and may be determined appropriately depending on the embodiment. The information processing device 1 may acquire one or more related words 6 for the search term 5 by accessing the meta-information database 60 using the search term 5 as a query.

(Related words)
The one or more related words 6 may include at least one of synonyms, similar words, higher-level concepts (hyperwords), and lower-level concepts (hyperwords) of the search term 5. A higher-level concept is a word that represents a general concept (higher concept) that includes the target word. For example, if the target word is CNG (Compressed Natural Gas), the higher-level concept may be fuel. A lower-level concept is a word that represents a specific concept (lower concept) included in the target word. For example, if the target word is fuel, the lower-level concept may be CNG.

This makes it possible to appropriately obtain related words 6 related to the search word 5. As a result, it is expected that the accuracy of searching sentences will be improved. Note that the related words 6 do not have to be limited to at least one of synonyms, similar words, higher-level conceptual words, and lower-level conceptual words. In another example,
When a target word is specified as a search term 5, any words desired to be used as additional search terms may be set as related terms of the target word.

(Categorization)
There are words (particularly abbreviations) that have the same notation but have different meanings. In addition, the meaning expressed by the target word may change depending on the situation in which it is used. To deal with this, the related words 6 may be categorized. Obtaining one or more related words 6 may consist of obtaining a category of the search word 5 and obtaining one or more related words 6 that belong to the obtained category. This makes it possible to appropriately obtain related words 6 that are highly related to the search word 5 in a search scene. As a result, it is expected that the search accuracy of sentences will be improved.

The method of specifying the category is not particularly limited and may be determined appropriately depending on the embodiment. In one example, the category may be specified by the user. In another example, the category may be associated with the document (plurality of sentences 2) to be searched. In this case, the target document is specified as the search target, and the category associated with the specified document may be obtained as the category of the search term 5.

The categories that can be specified may include a common category ("Common to all categories" in FIG. 5 described later) indicating that the category belongs to all categories. When a common category exists, obtaining one or more related words 6 that belong to the obtained category may consist of obtaining one or more related words 6 that belong to at least one of the category of the search term 5 and the common category.

In this embodiment, the related words 6 may be categorized and stored in the meta-information database 60. The meta-information database 60 may be appropriately configured to be able to acquire the related words 6 according to the specified word and category.

[2 Configuration Example]
[Hardware configuration example]
Fig. 2 is a schematic diagram showing an example of a hardware configuration of the information processing device 1 according to the present embodiment. As shown in Fig. 2, the information processing device 1 according to the present embodiment is a computer in which a control unit 11, a storage unit 12, an input device 13, an output device 14, and a drive 15 are electrically connected.

The control unit 11 includes a CPU (Central Processing Unit), which is a hardware processor;
The control unit 11 includes a random access memory (RAM), a read only memory (ROM), etc., and is configured to execute information processing based on programs and various data. The control unit 11 (CPU) is an example of a processor resource.

The storage unit 12 is configured, for example, with a hard disk drive, a solid state drive, or the like. The storage unit 12 is an example of a memory resource. In this embodiment, the storage unit 12 stores various information such as a search program 81, multiple documents 2, structural information 200, and a meta information database 60.

The search program 81 is a program for causing the information processing device 1 to execute a search process (see FIG. 4, described below) for extracting sentences that meet the search criteria. The search program 81 includes a series of instructions for the information processing.

The input device 13 is, for example, a device for inputting information such as a mouse or a keyboard. The output device 14 is, for example, a device for outputting information such as a display or a speaker. A user can operate the information processing device 1 by using the input device 13 and the output device 14. The input device 13 and the output device 14 may be integrally configured as, for example, a touch panel display or the like.

The drive 15 is a device for reading various information such as programs stored in the storage medium 91. At least one of the search program 81, the multiple sentences 2, the structural information 200, and the meta-information database 60 may be stored in the storage medium 91. In this case, the information processing device 1 may acquire at least one of the search program 81, the multiple sentences 2, the structural information 200, and the meta-information database 60 from the storage medium 91. The storage medium 91 is a medium that accumulates information such as programs by electrical, magnetic, optical, mechanical, or chemical action so that a computer or other device, machine, etc. can read the various information such as the stored programs.

In FIG. 2, a disk-type storage medium such as a CD or DVD is illustrated as an example of the storage medium 91. However, the type of the storage medium 91 is not limited to a disk type, and may be other than a disk type. An example of a storage medium other than a disk type is a semiconductor memory such as a flash memory. The type of the drive 15 may be appropriately selected depending on the type of the storage medium 91.

In addition, with regard to the specific hardware configuration of the information processing device 1, components can be omitted, replaced, or added as appropriate depending on the embodiment. For example, the control unit 11 may include multiple hardware processors. The hardware processor may be configured with a microprocessor, a field-programmable gate array (FPGA), a graphics processing unit (GPU), or the like. At least one of the input device 13, the output device 14, and the drive 15 may be omitted.

The information processing device 1 may further include a communication module for performing data communication with other computers. The communication module is, for example, a wired LAN (Local Area Network) module, a wireless LAN module, or the like. In this case, the information processing device 1 may use the communication module to perform data communication with other computers. For example, the information processing device 1 may receive a query including a search term 5 from the other computer, perform a search process according to the received query, and return the search results to the other computer. In addition, at least one of the multiple sentences 2, the structural information 200, and the meta-information database 60 may be stored in another computer such as a network server. The information processing device 1 may obtain at least one of the multiple sentences 2, the structural information 200, and the meta-information database 60 by accessing the other computer.

The information processing device 1 may further include an external interface for connecting to an external device. The external interface is, for example, a USB (Universal Serial Bus) port, a dedicated port, etc. In this case, the information processing device 1 may be connected to the external device using the external interface. For example, at least one of the search program 81, the multiple sentences 2, the structural information 200, and the meta information database 60 may be stored in an external storage device. The information processing device 1 may be connected to the external storage device via the external interface. The information processing device 1 may acquire at least one of the search program 81, the multiple sentences 2, the structural information 200, and the meta information database 60 from the external storage device.

The information processing device 1 may be configured by a plurality of computers. In this case, the hardware configurations of the computers may or may not be the same. The information processing device 1 may be a computer designed specifically for the service provided, a general-purpose server device, a general-purpose PC (Personal Computer), a mobile phone including a smartphone, a tablet PC, or the like.

[Software configuration example]
3 is a schematic diagram showing an example of a software configuration of the information processing device 1 according to the present embodiment. The control unit 11 of the information processing device 1 loads the search program 81 stored in the storage unit 12 in the RAM, and executes the instructions included in the search program 81 loaded in the RAM by the CPU. As a result, the information processing device 1 according to the present embodiment operates as a computer including a word acquisition unit 111, a word expansion unit 112, a structure acquisition unit 113, a structure identification unit 114, a range expansion unit 115, a search processing unit 116, and an output unit 117 as software modules. That is, in the present embodiment, each software module of the information processing device 1 is realized by the control unit 11 (CPU).

The word acquisition unit 111 is configured to acquire a search term 5. The word expansion unit 112 is configured to acquire one or more related words 6 related to the search term 5. The structure acquisition unit 113 is configured to acquire structure information 200 indicating a hierarchical relationship in a plurality of sentences 2 including a search target sentence 20, which are hierarchically organized sentences. The structure identification unit 114 is configured to identify an upper-level sentence 25 located in a higher level than the search target sentence 20 from among the plurality of sentences 2 by referring to the acquired structure information 200. The range expansion unit 115 is configured to acquire tag information 35 assigned to the identified upper-level sentence 25. In this embodiment, the range expansion unit 115 may be further configured to acquire tag information 30 assigned to the search target sentence 20. The search processing unit 116 is configured to search for whether or not a word matching the acquired search term 5 exists in the tag information 35 of the search target sentence 20 and the upper-level sentence 25. In this embodiment, the search processing unit 116 may be configured to search for whether or not there is a word matching at least one of the search term 5 and one or more related words 6 in the search target sentence 20, the tag information 30 of the search target sentence 20, and the tag information 35 of the upper level sentence 25. The output unit 117 is configured to output information related to the search result.

In this embodiment, an example is described in which each software module of the information processing device 1 is realized by a general-purpose CPU. However, some or all of the above software modules may be realized by one or more dedicated processors. Each of the above modules may be realized as a hardware module. Furthermore, with regard to the software configuration of the information processing device 1, modules may be omitted, replaced, or added as appropriate depending on the embodiment.

[3 Operation example]
4 is a flowchart showing an example of a processing procedure for text search by the information processing device 1 according to the present embodiment. The processing procedure described below is an example of a search method (information processing method) according to one aspect of the present disclosure. However, the processing procedure described below is merely an example, and each step may be changed as much as possible. In addition, steps may be omitted, replaced, or added to the following processing procedure as appropriate depending on the embodiment.

(Step S101)
In step S101, the control unit 11 operates as the word acquisition unit 111 and acquires the search term 5. The manner in which the search term 5 is acquired may be appropriately selected depending on the embodiment. In one example, the control unit 11 may acquire the search term 5 in response to a user's input via the input device 13. In another example, the control unit 11 may acquire the search term 5 from an arbitrary sentence. In yet another example, the control unit 11 may acquire the search term 5 from another computer. After acquiring the search term 5, the control unit 11 proceeds to the next step S102.

(Step S102)
In step S<b>102 , the control unit 11 operates as the word expansion unit 112 and acquires one or more related words 6 related to the search word 5 .

The method of acquiring the related words 6 may be appropriately determined according to the embodiment. In this embodiment, the one or more related words 6 may include at least one of synonyms, synonyms, higher-level conceptual terms, and lower-level conceptual terms of the search term 5. The related words 6 may also be categorized. In this case, the control unit 11 may acquire the category of the search term 5. In one example, the control unit 11 may acquire the category in response to a user's input via the input device 13. In another example, the control unit 11 may acquire a category associated with the specified document as a category of the search term 5 by specifying the target document as a search target. In yet another example, the control unit 11 may acquire a category from another computer. The acquisition of the category may be performed at any timing before the processing of step S102 is performed. For example, the control unit 11 may execute a process of acquiring a category in parallel with the processing of step S101. Then, the control unit 11 may acquire one or more related words 6 belonging to the acquired category. In this embodiment, the control unit 11 can acquire one or more related words 6 from the meta-information database 60 by using the search term 5 as a query.

FIG. 5 is a schematic diagram showing an example of a meta-information database 60 according to the present embodiment. The meta-information database 60 shown in FIG. 5 has a data structure in the form of a table. Each record (row data) has fields for storing the values of a target word, a category, a type, and a related word. The target word field stores a word to be searched for related words. The category field stores a category to which the related word belongs. In the example of FIG. 5, "common to all categories" indicates that the related word belongs to all categories. The type field stores information indicating the type of related word (in the example of FIG. 5, synonymous word, similar word, higher-level word, or lower-level word). The related word field stores related words related to the target word. As shown in FIG. 5, a target word may have multiple related words of the same type. Accordingly, each record may have multiple related word fields ("first related word" and "second related word" in FIG. 5).

The control unit 11 may use the acquired search term 5 and category as a query to perform a search process for related terms 6 in the meta information database 60. In one example of the search process, the control unit 11 may refer to the target word field and category field of each record, and extract records in which the same values as the acquired search term 5 and category are stored in the target word field and category field. If a common category exists, the control unit 11 may further extract records in which the same value as the acquired search term 5 is stored in the target word field, and a value indicating the common category is stored in the category field. The control unit 11 may then acquire the related terms 6 from the related word field of the extracted records.

When the related words 6 include a higher-level concept word, the control unit 11 may recursively acquire a higher-level concept word of the search word 5. That is, the control unit 11 may use the higher-level concept word and the category of the search word 5 as a query to further acquire a related word of the higher-level concept word from the meta-information database 60 as a related word 6 of the search word 5. The control unit 11 may acquire a higher-level concept word and a related word of the higher-level concept word as a related word 6 of the search word 5 by recursively repeating the acquisition process of the related words 6 until the related word of the highest-level concept word is acquired. The same may be true when the related words 6 include a lower-level concept word. That is, the control unit 11 may use the lower-level concept word and the category of the search word 5 as a query to further acquire a related word of the lower-level concept word from the meta-information database 60 as a related word 6 of the search word 5. The control unit 11 may acquire a lower-level concept word and a related word of the lower-level concept word as a related word 6 of the search word 5 by recursively repeating the acquisition process of the related words 6 until the related word of the lowest-level concept word is acquired.

Note that, as long as it is possible to acquire related words 6 corresponding to the specified search term 5, the data structure of the meta information database 60 does not need to be limited to the example shown in FIG. 5, and may be appropriately determined depending on the embodiment. When one or more related words 6 are acquired, the control unit 11 proceeds to the next step S103.

(Steps S103 to S105)
4, in step S103, the control unit 11 operates as the structure acquisition unit 113 and acquires the structure information 200. After acquiring the structure information 200, the control unit 11 advances the process to the next step S104.

In step S104, the control unit 11 operates as a structure identification unit 114 and identifies, from among the multiple sentences 2, a higher-level sentence 25 that is located in a higher level than the search target sentence 20 by referring to the acquired structure information 200. If multiple higher-level sentences 25 exist, the control unit 11 may trace the hierarchy upward from the search target sentence 20 and identify at least a portion of the multiple higher-level sentences 25.

The search target sentence 20 may be arbitrarily selected from among a plurality of sentences 2. In one example, the control unit 11 may acquire the search target sentence 20 in response to a user selection via the input device 13. In another example, the control unit 11 may acquire the search target sentence 20 in an arbitrary manner from among a plurality of sentences 2 designated as search targets. The plurality of sentences 2 to be searched may be designated at any timing before the process of step S103 is executed. The search target sentence 20 may be selected at any timing before the process of step S104 is executed. When the upper-level sentence 25 is identified, the control unit 11 proceeds to the next step S105.

In step S105, the control unit 11 operates as the range expansion unit 115 and acquires the tag information 35 assigned to the identified upper-level sentence 25. In this embodiment, the control unit 11 may further acquire the tag information 30 assigned to the search target sentence 20. If there are multiple upper-level sentences 25, the control unit 11 may acquire the tag information 35 of at least a portion of the multiple upper-level sentences 25. In one example, the control unit 11 may go back one layer at a time from the search target sentence 20 to the top upper-level sentence 25 and acquire the tag information 35 of the upper-level sentence 25 of each layer. In this way, the control unit 11 expands the matching range of the search target sentence 20.

Figure 6 shows a schematic example of structural information 200 and hierarchical sentences according to this embodiment. Figure 7 shows a schematic example of the matching range of each sentence after a process is executed to obtain tag information of a higher-level sentence for each sentence shown in Figure 6. Note that in the examples of Figures 6 and 7, the structural information 200, each sentence, and tag information are expressed in table format for ease of explanation. However, the data format of the structural information 200, each sentence, and tag information is not limited to this example, and may be determined appropriately depending on the embodiment.

In the example of FIG. 6, it is assumed that each document is divided into three hierarchical levels by chapter. The document "Attachment 42" is the top level, "Attachment 42 I" is the middle level, and "Attachment 42 I" is the second level.
"1." is at the lowest level. The text of "Attachment 42" is given the "first hierarchical tag information" in FIG. 6. The text of "Attachment 42 I" is given the "second hierarchical tag information."
The sentence "Attachment 42 I 1." is given "third-level tag information." When the sentence "Attachment 42 I" is the search target sentence 20, the higher-level sentence 25 is the sentence "Attachment 42." When the sentence "Attachment 42 I 1." is the search target sentence 20, the higher-level sentences 25 are the sentences "Attachment 42 I" and "Attachment 42."

As shown in Fig. 7, the control unit 11 may obtain tag information of the sentence "Attachment 42 I" and the sentence "Attachment 42" for the sentence "Attachment 42 I". As a result, the control unit 11 may expand the collation range of the sentence "Attachment 42 I" to the sentence "Attachment 42 I", the tag information of "Attachment 42", and the tag information of "Attachment 42 I". In addition, the control unit 11 may obtain tag information of the sentence "Attachment 42 I 1.", the sentence "Attachment 42 I", and the sentence "Attachment 42" for the sentence "Attachment 42 I 1.". As a result, the control unit 11 may expand the collation range of the sentence "Attachment 42 I" to the sentence "Attachment 42 I", the tag information of "Attachment 42", and the tag information of "Attachment 42 I".
The matching range of the sentence "Attachment 42 I 1." may be expanded to the sentence "Attachment 42 I 1.", the tag information of "Attachment 42", the tag information of "Attachment 42 I", and the tag information of "Attachment 42 I 1.".
7, the range of matching for the text "Attachment 42 I 1." is indicated by a bold frame. After acquiring the tag information (30, 35), the control unit 11 advances the process to the next step S106.

(Step S106)
In step S106, the control unit 11 searches the tag information 35 of the search target sentence 20 and the upper hierarchical sentence 25 to determine whether or not there is a word matching the acquired search term 5. In this embodiment, the control unit 11 may search the search target sentence 20, the tag information 30 of the search target sentence 20, and the tag information 35 of the upper hierarchical sentence 25 to determine whether or not there is a word matching at least one of the search term 5 and one or more related words 6.

When the search term 5 and related word 6 are used as conditions for a logical sum search, in the search process, the control unit 11 determines whether or not a word matching the search term 5 or related word 6 exists within the matching range of the search target sentence 20. If a word matching the search term 5 or related word 6 exists within the matching range, the control unit 11 extracts the search target sentence 20 as the search result. On the other hand, if a word matching the search term 5 or related word 6 does not exist within the matching range, the control unit 11 does not extract the search target sentence 20.

When there are multiple search target sentences 20, the control unit 11 may execute at least the processing of steps S104 to S106 for each search target sentence 20. For example, the control unit 11 may select a search target sentence to be processed from the multiple search target sentences 20. The control unit 11 may execute the processing of steps S104 to S106 for the selected search target sentence. The selection method may be set arbitrarily (for example, the selection order may be specified in advance). The control unit 11 may repeatedly execute this series of information processing until the processing of steps S104 to S106 is completed for all search target sentences 20. The processing of step S103 may be executed appropriately. In one example, the processing of step S103 may be executed for all search target sentences 20 at once. In another example, the processing of step S103 may be executed for each search target sentence 20.

As a specific example, assume that various information in Figs. 5 to 7 is given, and that in the process of step S101, two search terms 5 "measurement method" and "fuel" are acquired as logical product conditions. Also assume that "exhaust gas regulations" is acquired as a category. In this case, in step S102, the control unit 11 may refer to the meta information database 60 to acquire related terms 6 for "measurement method" and "fuel". In the example of Fig. 5, since there is no record that stores "measurement method" in the target word field, related terms 6 for "measurement method" are not obtained. On the other hand, there is a record that stores "fuel" in the target word field. The control unit 11 can acquire "CNG", "compressed natural gas", "natural gas", "LPG", and "liquefied petroleum gas" as related terms 6 for "fuel" by recursively acquiring subordinate concepts under the condition that "exhaust gas regulations" is specified as the category. It is assumed that these related terms 6 are used as logical sum conditions.

7, when the search target sentence 20 is the sentence of "Attachment 42 I 1.", the control unit 11 may acquire tag information of "Attachment 42", tag information of "Attachment 42 I", and tag information of "Attachment 42 I 1." by the processing of steps S103 to S105. In the processing of step S106, the control unit 11 may search whether or not there is a word matching the above-mentioned search term 5 or related word 6 in the sentence of "Attachment 42 I 1.", tag information of "Attachment 42", tag information of "Attachment 42 I", and tag information of "Attachment 42 I 1.". "Measurement method" is present in the tag information of "Attachment 42". "Fuel", "CNG", "Compressed natural gas", "Natural gas", "LPG", and "Liquefied petroleum gas" are present in the sentence and tag information of "Attachment 42 I 1.". Therefore, the control unit 11 can extract the sentence "Attachment 42 I 1." as a sentence that meets the search conditions of "measurement method" and "fuel". When the search process is completed, the control unit 11 proceeds to the next step S107.

(Step S107)
In step S107, the control unit 11 operates as the output unit 117 and outputs information related to the search results.

The output destination and the contents of the output process may be appropriately determined according to the embodiment. The output destination may be selected from, for example, RAM, the storage unit 12, the output device 14, the storage medium 91, an external storage device, another computer, etc. In one example, the control unit 11 may output a list of sentences extracted as search results to the output device 14 (for example, a display). In this case, the control unit 11 may display the search term 5 and related words 6 used as search conditions together with the extracted sentences. In addition, the control unit 11 may clearly indicate the positions of the search term 5 and related words 6 used in the extracted sentences. In another example, the control unit 11 may transmit the search results to another computer. The search results may be displayed in any manner (for example, the same manner as described above) in the other computer.

When the output of the information is completed, the control unit 11 ends the processing procedure of the information processing device 1 according to this operation example. The control unit 11 may repeatedly execute the processes of steps S101 to S107. The timing for repeating the processes may be appropriately determined according to the embodiment. As a result, the information processing device 1 may repeatedly execute the search process for sentences matching the search term 5.

(Features)
In this embodiment, the process of steps S103 to S105 expands the matching range of the search term 5 to not only the search target sentence 20 but also the tag information 30 of the search target sentence 20 and the tag information 35 of the upper layer sentence 25. As a result, in the process of step S106, even if the search target sentence 20 does not contain a word matching the search term 5, if the tag information (30, 35) contains a word matching the search term 5, the search target sentence 20 can be extracted as a search result. For example, in the above specific example, "measurement method" is not included in the sentence "Attachment 42 I 1.", but is included in the tag information of "Attachment 42.". Therefore, the sentence "Attachment 42 I 1." can be extracted as a search result. In addition, in this embodiment, this tag information 35 is added to the matching range instead of the upper layer sentence 25 itself. This makes it easier to appropriately extract the search target sentence 20 while suppressing the amount of expansion of the matching range. Therefore, according to this embodiment, the searchability of the sentence can be efficiently improved.

[4 Modifications]
Although the embodiment of the present disclosure has been described in detail above, the above description is merely an example of the present disclosure in every respect. It goes without saying that various improvements or modifications can be made without departing from the scope of the present disclosure. For example, the following modifications are possible. The following modifications can be combined as appropriate.

In the above embodiment, the control unit 11 may omit the process of acquiring tag information 30 assigned to the search target sentence 20.

In the above embodiment, the related words 6 do not need to be categorized. In this case, each process related to categories may be omitted in the above processing procedure. The category field may be omitted from each record in the meta information database 60.

In the above embodiment, the process of acquiring related words 6 may be omitted. In this case, the word expansion unit 112 may be omitted from the software configuration of the information processing device 1. In the above processing procedure, the process of step S102 may be omitted. The meta information database 60 may be omitted.

In the processing procedure (FIG. 4) according to the above embodiment, the order of processing of each step may be changed as appropriate. For example, the processing related to the search term 5 in steps S101 and S102 may be performed at any time before the processing of step S106 is performed. The processing related to the expansion of the search target sentence 20 in steps S103 to S105 may be performed at any time before the processing of step S106 is performed. The processing related to the search term 5 in steps S101 and S102 may be performed at least partially in parallel with the processing related to the expansion of the search target sentence 20 in steps S103 to S105. The search target sentence 20 may be specified at any time before the processing of step S103 or step S104 is performed.

[5 Supplementary Information]
The processes and means described in the present disclosure can be freely combined as long as no technical contradiction occurs.

In addition, a process described as being performed by one device may be shared and executed by multiple devices. Or, a process described as being performed by different devices may be executed by a single device. In a computer system, the hardware configuration that realizes each function can be flexibly changed.

The present disclosure can also be realized by supplying a computer program that implements the functions described in the above embodiments to a computer, and having one or more processors of the computer read and execute the program. Such a computer program may be provided to the computer by a non-transitory computer-readable storage medium that can be connected to the system bus of the computer, or may be provided to the computer via a network. Non-transitory computer-readable storage media include, for example, any type of disk, such as a magnetic disk (floppy disk, hard disk drive (HDD), etc.), an optical disk (CD-ROM, DVD disk, Blu-ray disk, etc.), a read-only memory (ROM), a random access memory (RAM), an EPROM, an EEPROM, a magnetic card, a flash memory, an optical card, or any type of medium suitable for storing electronic instructions.

1... Information processing device,
11: control unit, 12: storage unit,
13...input device, 14...output device, 15...drive,
81...search program, 91...storage medium,
111...word acquisition unit, 112...word expansion unit,
113...Structure acquisition unit, 114...Structure identification unit,
115: range expansion unit, 116: search processing unit,
117...output unit,
2. Multiple sentences,
20...search target sentence, 25...upper-level sentence,
200...Structural information,
30...tag information (added to the search target text),
35...Tag information (assigned to upper-level documents),
5...search term, 6...related words,
60…Metadata database

Claims (1)

1. A computer-implemented search method, comprising:
Obtaining search terms;
Obtaining structural information indicating a hierarchical relationship among a plurality of hierarchically organized sentences including the search target sentence;
by referring to the acquired structure information, identifying, from among the plurality of sentences, a higher-level sentence that is located in a higher level than the sentence to be searched;
acquiring tag information assigned to the identified higher-level document;
searching for a word that matches the acquired search term in tag information of the search target sentence and the upper level sentence; and outputting information related to the search result.
Including,
How to search.

Images