Computer system and data search method

Abstract

To realize a system for asking questions for efficiently searching for data.SOLUTION: A computer system is connected to a database that stores a plurality of data. The data is assigned a plurality of tags to be used in search, and the tag is assigned either a first attribute value indicating that the data matches an event related to the tag, or a second attribute value indicating that the data does not match the event related to the tag. When receiving a search request, the computer system obtains an attribute value of a tag from the search request, analyzes a relation between the attribute value of the tag of the data and the obtained attribute value of the tag, refers to the database based on analysis results, searches for data to be output as a response to the search request, generates a question sentence for confirming the attribute value of the tag based on the analysis results, and outputs at least one of the searched data and the question sentence.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a technology for supporting data searches. [Background technology]

[0002] Tag search is a well-known data search technique. By attaching tags to data, data can be searched efficiently and quickly. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-187109 Summary of the Invention [Problem to be solved by the invention]

[0004] If the tag information included in a search request is insufficient, the desired data cannot be searched for. Therefore, additional tag information is required to identify the data. For example, a search system may ask a question to acquire tag information. The technology described in Patent Document 1 is known as a technology for generating questions.

[0005] Patent Document 1 states that "the dialogue knowledge generation device comprises a keyword selection unit that selects attribute values, which are dialogue keywords, and attribute names, which indicate categories of attribute values, from target data including business documents and past dialogue records using clue phrases that indicate the respective selection conditions; a correspondence unit that determines the correspondence between the attribute values ​​and attribute names based on their respective acquisition positions in the target data; and a graph generation unit that extracts the hierarchical relationship of multiple items having attribute values ​​or attribute names, and generates a knowledge graph by combining the multiple items based on the extracted hierarchical relationship."

[0006] The present invention aims to provide a system and method for querying data efficiently. [Means for solving the problem]

[0007] A representative example of the invention disclosed in the present application is as follows: That is, a computer system is connected to a database in an accessible manner for storing a plurality of data, the data being assigned a plurality of tags used for searching, the tags being assigned either a first attribute value indicating that the data corresponds to a matter related to the tag or a second attribute value indicating that the data does not correspond to a matter related to the tag, when the computer system receives a search request including information indicating whether or not at least one of the tags corresponds to a matter related to the tag, the computer system acquires the attribute values ​​of the plurality of tags from the search request, analyzes the relationship between the attribute values ​​of the plurality of tags of the data and the acquired attribute values ​​of the plurality of tags, references the database based on the result of the analysis, searches for the data to be output as a response to the search request, generates a question to confirm the attribute values ​​of the tags based on the result of the analysis, and outputs at least one of the searched data and the question. [Effects of the Invention]

[0008] According to the present invention, queries can be made to efficiently search for data. Objects, configurations and effects other than those described above will become apparent from the following description of the embodiments. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 illustrates an example of a system configuration according to a first embodiment. [Figure 2] FIG. 2 is a diagram illustrating an example of a hardware configuration of a computer that configures the search system according to the first embodiment. [Figure 3] FIG. 2 is a diagram illustrating an example of a database according to the first embodiment. [Figure 4] FIG. 10 is a diagram illustrating an example of generation policy information according to the first embodiment. [Figure 5] FIG. 10 is a diagram illustrating an example of category information according to the first embodiment. [Figure 6] FIG. 10 is a diagram illustrating an example of correlation information according to the first embodiment. [Figure 7] 10 is a flowchart illustrating an example of a search process executed by the search system according to the first embodiment. [Figure 8] 10 is a flowchart illustrating an example of a tag analysis process executed by the search system according to the first embodiment. [Figure 9] FIG. 10 is a diagram showing an example of attribute values ​​of each tag acquired from a search request in the first embodiment. [Figure 10] FIG. 10 is a diagram showing an example of analysis information according to the first embodiment. [Figure 11] 10 is a flowchart illustrating an example of a question generation process executed by the search system according to the first embodiment. [Figure 12] FIG. 10 is a diagram illustrating an example of a calculation process of a search unit according to the first embodiment. [Figure 13] FIG. 10 is a diagram illustrating an example of a database according to a second embodiment. [Figure 14] FIG. 11 is a diagram illustrating an example of generation policy information according to the second embodiment. [Figure 15] FIG. 10 is a diagram showing an example of analysis information according to the second embodiment. [Figure 16] FIG. 10 is a diagram illustrating an example of calculation processing of a search unit according to the second embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0010] Hereinafter, embodiments of the present invention will be described with reference to the drawings. However, the present invention should not be construed as being limited to the description of the embodiments shown below. Those skilled in the art will readily understand that the specific configuration can be changed without departing from the spirit or intent of the present invention.

[0011] In the configuration of the invention described below, the same or similar configurations or functions are denoted by the same reference numerals, and redundant explanations will be omitted.

[0012] In this specification, the terms "first," "second," "third," etc. are used to identify components and do not necessarily limit the number or order. [Example]

[0013] Fig. 1 is a diagram illustrating an example of a system configuration of Example 1. Fig. 2 is a diagram illustrating an example of a hardware configuration of a computer that configures the search system of Example 1.

[0014] The system is composed of a search system 100 and a terminal 101. The search system 100 and the terminal 101 are connected directly or via a network such as a LAN (Local Area Network).

[0015] Terminal 101 is a terminal operated by a user who searches for data. Search system 100 manages data and searches for data. A user operates terminal 101 to send a search request to search system 100. When search system 100 receives a search request, it searches for data. When information to identify data is insufficient, search system 100 generates a question and sends it to terminal 101. When data matching the search conditions is found, search system 100 sends the data to terminal 101. A user sends an answer to the question to search system 100 via terminal 101. When search system 100 receives an answer, it searches for data based on the search request and the answer.

[0016] The search system 100 is configured, for example, from a computer 200 as shown in Fig. 2. The computer 200 has a processor 201, a network interface 202, a main memory device 203, and a secondary memory device 204. Each hardware element is connected via a bus 205.

[0017] The processor 201 executes a program stored in the main memory device 203. The processor 201 executes processing in accordance with the program, thereby operating as a functional unit (module) that realizes a specific function. In the following explanation, when a processing is explained using a functional unit as the subject, it indicates that the processor 201 is executing a program that realizes the functional unit.

[0018] The network interface 202 communicates with external devices via a network. The main memory device 203 stores programs and information executed by the processor 201. The main memory device 203 is also used as a work area. The secondary memory device 204 is a hard disk drive (HDD) or solid state drive (Solid State Drive) or the like.

[0019] The search system 100 includes a data management unit 110 and a search unit 111. The search system 100 also holds a database 120, creation policy information 121, category information 122, and correlation information 123.

[0020] The data management unit 110 manages the database 120. The search unit 111 searches for data stored in the database 120.

[0021] The database 120 stores data to be searched. The data may be documents, images, etc. As will be described later, tags for search purposes are attached to the data. The present invention is not limited to the type of data stored in the database 120.

[0022] The data management unit 110 and the database 120 may be implemented as separate systems from the search unit 111, the generation policy information 121, the category information 122, and the correlation information 123.

[0023] FIG. 3 is a diagram illustrating an example of the database 120 according to the first embodiment.

[0024] The database 120 stores a table 300. The table 300 stores entries each including a name 301, a content 302, and tag information 303. There is one entry for one piece of data.

[0025] Name 301 is a field for storing the name of the data. Content 302 is a field for storing the content of the data.

[0026] The tag information 303 is a field that stores information about tags assigned to data. In this embodiment, the same tag is assigned to all data, and the tags are managed in association with attribute values ​​that indicate whether the tag-related matter applies. The tag information 303 has as many fields as there are tags, and each field stores an attribute value. A category is also set for the tag. Each field of the tag information 303 stores either an attribute value "1" that indicates that the tag-related matter applies, or an attribute value "0" that indicates that the tag-related matter does not apply.

[0027] For example, when data on accommodations is stored in database 120, possible categories include "purpose," "room type," "facilities," and "price range," and possible tags include "business trip," "double," "hot spring," and "high class." If the attribute value of the tag "business trip" in the category "purpose" is "1," it indicates that the room is a business trip. Also, if the attribute value of the tag "double" in the category "room type" is "0," it indicates that the room does not correspond to a double room.

[0028] FIG. 4 is a diagram illustrating an example of the generation policy information 121 according to the first embodiment.

[0029] The generation policy information 121 is information for managing the generation policy of a question sentence. The generation policy information 121 stores entries including a relationship ID 401, an attribute value (data) 402, an attribute value (user input) 403, and a generation policy 404. There is one entry for each combination of the attribute value of a tag in the data and the attribute value of a tag in the user input.

[0030] The relationship ID 401 is a field for storing an ID for identifying the relationship between the attribute value of a tag of data and the attribute value of a tag in a user input. In the following description, the relationship between the attribute value of a tag of data and the attribute value of a tag in a user input is simply referred to as a relationship.

[0031] Attribute value (data) 402 is a field that stores the attribute value of the tag of the data. Attribute value (user input) 403 is a field that stores the attribute value of the tag in the user input. "-" indicates that the tag attribute value is not included in the user input.

[0032] Generation policy 404 is a field for storing the question generation policy. "Not required" indicates that the generation of a question is not required. "Tag confirmation" is a generation policy for generating a question for confirming the attribute value of a target tag. For example, a possible question would be "Does this correspond to matters related to tag X?". "Tag change" is a generation policy for generating a question for confirming the attribute value of another tag that is related to the target tag. For example, a possible question would be "Does this correspond to matters related to tag Y in the same category as tag X?". Note that with either generation policy, a question for confirming the attribute value of a tag will be generated.

[0033] "Content confirmation; tag confirmation" is a generation policy that generates a question to prompt the user to confirm the content of the data and, if necessary, to confirm the attribute value of the tag entered by the user. "Content confirmation; tag change" is a generation policy that generates a question to prompt the user to confirm the content of the data and, if necessary, to change the attribute value of the tag entered by the user. The "Data content confirmation" for each generation policy is described below.

[0034] When a tag's attribute value is "1," it is unclear whether the search system should only consider cases that match the tag's criteria, or whether it should consider both cases that match and cases that do not. Therefore, the search system 100 prompts the user to confirm the data content. When confirming the data content, in the former case, where only cases that match the tag's criteria are considered, a question is generated to confirm the tag's attribute value entered by the user, or a question is generated to change the tag's attribute value entered by the user. In the latter case, where both cases that match and cases that do not match the tag's criteria are considered, no question generation is required.

[0035] FIG. 5 is a diagram illustrating an example of the category information 122 according to the first embodiment.

[0036] The category information 122 is information for managing tag categories. The category information 122 stores entries including a tag name 501 and a category 502. There is one entry for one tag. The category 502 is a field for storing the tag category.

[0037] FIG. 6 is a diagram illustrating an example of the correlation information 123 according to the first embodiment.

[0038] The correlation information 123 is information for managing correlations between tags. The correlation information 123 is information in a matrix format, and correlation values ​​between tags are stored in cells.

[0039] FIG. 7 is a flowchart illustrating an example of a search process executed by the search system 100 of the first embodiment. FIG. 8 is a flowchart illustrating an example of a tag analysis process executed by the search system 100 of the first embodiment. FIG. 9 is a diagram illustrating an example of an attribute value of each tag acquired from a search request of the first embodiment. FIG. 10 is a diagram illustrating an example of analysis information of the first embodiment. FIG. 11 is a flowchart illustrating an example of a question generation process executed by the search system 100 of the first embodiment. FIG. 12 is a diagram illustrating an example of a calculation process of the search unit 111 of the first embodiment.

[0040] When the search unit 111 receives a search request from the terminal 101, it starts a search process.

[0041] The search unit 111 acquires the attribute value of each tag from the search request (S101). The search request includes, for at least one tag, information indicating whether a matter related to the tag applies. The information is, for example, text. Note that the attribute value of a tag that does not include information indicating whether a matter related to the tag applies is "-". The attribute value of a tag can be acquired using a known technique such as text analysis.

[0042] Next, the search unit 111 executes tag analysis processing (S102). Here, the tag analysis processing will be described with reference to FIG.

[0043] First, the search unit 111 initializes the analytical information 1000 (S201). As shown in FIG. 10, the analytical information 1000 is information in a matrix format with data in rows and tags in columns. A relationship ID is stored in each cell. In S201, analytical information 1000 is generated in which all cells are blank.

[0044] The search unit 111 starts loop processing of the data (S202). Here, the search unit 111 selects one row (data) of the analysis information 1000.

[0045] The search unit 111 starts a loop process of tags (S203). Here, the search unit 111 selects one column (tag) of the analysis information 1000.

[0046] The search unit 111 refers to the generation policy information 121 to identify the relationship between the attribute value of the selected tag of the selected data and the attribute value of the selected tag obtained from the search request, and records it in the corresponding cell of the analysis information 1000 (S204).

[0047] The search unit 111 determines whether or not the process has been completed for all tags (S205). If the process has not been completed for all tags, the search unit 111 returns to S203.

[0048] When the processing is completed for all tags, the search unit 111 determines whether the processing is completed for all data (S206). When the processing is not completed for all data, the search unit 111 returns to S202. When the processing is completed for all data, the search unit 111 ends the tag analysis processing.

[0049] When the attribute values ​​of each tag acquired from the search request are as shown in FIG. 9, analysis information 1000 shown in FIG. 10 is generated.

[0050] This concludes the description of the tag analysis process. Returning to the description of Fig. 7, the search unit 111 searches for data based on the attribute values ​​of each tag acquired from the search request, based on the analysis information 1000 (S103). For example, the search unit 111 searches for data that does not have a relationship "D" as data that matches the search criteria specified in the search request.

[0051] The search unit 111 executes a question generation process (S104). Here, the question generation process will be described with reference to FIG.

[0052] The search unit 111 selects a target tag based on the analysis information 1000 (S301). Specifically, the following process is executed.

[0053] (S301-1) The search unit 111 refers to the analysis information 1000 and calculates, for each tag, the number of data pieces for which a relationship is established (first value). For example, the information shown in the upper part of FIG. 12 is generated.

[0054] (S301-2) The search unit 111 calculates the sum of the first values ​​of each tag for each question generation policy. For example, if the question generation policy is "tag check," the search unit 111 calculates the sum of the first values ​​of the relationship "E" for each tag. This process generates the information shown in the lower part of FIG. 12.

[0055] In this way, by asking questions that confirm the attribute values ​​of tags with which a large number of data items have relationships, it is expected that data can be narrowed down efficiently.

[0056] (S301-3) Based on the result of S301-2, the search unit 111 selects target tags for each question generation policy. For example, the search unit 111 selects a predetermined number of tags in descending order of total. The search unit 111 also selects tags whose total is greater than a threshold.

[0057] The search unit 111 starts a loop process of question generation policies (S302), where the search unit 111 selects one question generation policy.

[0058] The search unit 111 starts a loop process of the target tags in the selected question generation policy (S303), where the search unit 111 selects one target tag from the target tags.

[0059] The search unit 111 determines the type of question (S304). The types of questions include a question for expanding the search range (expanding question) and a question for narrowing the search range (narrowing question).

[0060] In this embodiment, the generation policy, the name of the target tag, and the attribute value of the target tag acquired from the search request are received as input, and the question type is determined using a machine learning model that outputs the question type.

[0061] The search unit 111 generates a question sentence based on the target tag, the attribute value of the target tag acquired from the search request, and the type of question (S305). The search unit 111 stores the target tag, the question sentence generation policy, and the question sentence in association with each other.

[0062] A possible narrowing question is, for example, a question for checking whether or not the subject corresponds to a matter related to the target tag. Also, a possible broadening question is, for example, a question for checking whether or not the subject corresponds to a matter related to a tag in the same category as the target tag. When generating the broadening question described above, the search unit 111 refers to the category information 122 and the correlation information 123.

[0063] The question may be generated using a dedicated algorithm or a machine learning model such as a large-scale language model.

[0064] If the question generation policy is "content confirmation; tag confirmation" or "content confirmation; tag change," a question is generated to confirm the content of the data. If necessary, a question is also generated to confirm the attribute value of the tag entered by the user.

[0065] The search unit 111 determines whether or not the process has been completed for all tags (S306). If the process has not been completed for all tags, the search unit 111 returns to S303.

[0066] When the process has been completed for all tags, the search unit 111 selects a question sentence (S307). Specifically, the following process is executed.

[0067] (S307-1) The search unit 111 determines the rank of the question. For example, the following rank determination method is possible.

[0068] (Determination Method 1) The search unit 111 refers to the correlation information 123, calculates the sum of correlation values ​​between tags acquired from the search request whose attribute value is "0" or "1" and target tags associated with the question, and determines the rank of the question based on the sum of correlation values. For example, the rank is determined in descending order of the sum of correlation values.

[0069] (Determination Method 2) The search unit 111 determines the rank of a question sentence based on the sum of the first values ​​for each question sentence generation policy shown in the lower part of Fig. 12. The larger the sum, the greater the number of data items for which tag attribute values ​​can be confirmed in one question sentence. Therefore, by preferentially outputting questions with large sums, efficient search becomes possible.

[0070] (Determination Method 3) The search unit 111 determines the rank of the question sentence based on the order in which the question sentences were created.

[0071] (S307-2) The search unit 111 selects questions based on their ranks. For example, the search unit 111 selects a predetermined number of questions in order of rank.

[0072] The search unit 111 determines whether or not the process has been completed for all question generation policies (S308). If the process has not been completed for all question generation policies, the search unit 111 returns to S302.

[0073] When the process has been completed for all question generation policies, the search unit 111 starts loop processing of the questions (S309). Here, the search unit 111 selects one question from the questions selected in S307.

[0074] The search unit 111 uses the selected question to identify data for checking the attribute value of the tag (S310).

[0075] Specifically, the search unit 111 refers to the analysis information 1000 and identifies data in which the relationship of the target tag associated with the selected question in the data matches the relationship corresponding to the question generation policy associated with the selected question. Note that the data searched in S103 is excluded.

[0076] The search unit 111 selects, from the identified data, candidate data for checking the attribute value of the tag based on the answer to the selected question (S311). Specifically, the following process is executed.

[0077] (S311-1) The search unit 111 calculates an index for evaluating the cost required to search the data identified in S310. For example, the following indexes can be considered.

[0078] (Index 1) The search unit 111 refers to the analysis information 1000 and calculates the number of tags for which either the relationship "B" or "C" holds for each data as an index. If there are many tags whose data content needs to be checked, the cost required for searching will be high.

[0079] (Index 2) The search unit 111 refers to the analysis information 1000 and calculates the number of tags for which either the relationship "C" or "D" holds for each piece of data as an index. If there are many tags to be changed, the number of new tags to be checked will increase, and the cost required for searching will increase.

[0080] (Index 3) The search unit 111 refers to the analysis information 1000 and calculates, as an index, the number of tags for which either the relationship "B" or "E" is established for each piece of data. If there are many tags to be checked, the cost required for the search will be high.

[0081] (Index 4) The search unit 111 refers to the analysis information 1000 and calculates, as an index, the number of tags for which either the relationship "A" or "F" holds for each piece of data. If there are fewer tags to check, the cost required for the search will be smaller.

[0082] (S311-2) The search unit 111 selects candidate data based on the index and registers the identification information of the data in a candidate data list. For indexes (index 1), (index 2), and (index 3), the search unit 111 selects a predetermined number of data in ascending order of index. For index (index 4), the search unit 111 selects a predetermined number of data in descending order of index.

[0083] The search unit 111 determines whether or not the process has been completed for all question sentences (S312). If the process has not been completed for all question sentences, the search unit 111 returns to S309. If the process has been completed for all question sentences, the search unit 111 ends the question sentence generation process.

[0084] This concludes the explanation of the question generation process. Returning to the explanation of Fig. 7, the search unit 111 transmits the searched data and the question to the terminal 101 (S105).

[0085] The search unit 111 determines whether to end the search (S106). For example, if a question sentence is not generated or if an end instruction is received from the user, the search unit 111 ends the search process.

[0086] If the search is not to be ended, the search unit 111 determines whether or not a response has been received from the user (S107).

[0087] If no answer has been received from the user, the search unit 111 returns to S107 and waits for a response to be received.

[0088] When an answer is received from the user, the search unit 111 acquires the attribute value of each tag from the answer (S108), after which the search unit 111 returns to S102.

[0089] When a response is received, the tag analysis process is partially different. Specifically, the process of S201 is not executed. Also, in the data loop process, the process is executed for the candidate data.

[0090] The search system 100 of the first embodiment can identify tags for which data attribute values ​​need to be checked, and generate a query to check the attribute value of the tag entered by the user as needed. This allows for efficient search of desired data. Furthermore, for tags with mismatched attribute values, the search system 100 of the first embodiment can widen the search targets in the database 120 by generating a query to check the attribute values ​​of other tags, thereby realizing a comprehensive search. [Example]

[0091] The second embodiment differs from the first embodiment in that the attribute value in the tag is expanded. The second embodiment will be described below, focusing on the differences from the first embodiment.

[0092] The system configuration of the second embodiment is the same as that of the first embodiment. The hardware configuration and software configuration of the search system 100 of the second embodiment are the same as those of the first embodiment. The category information 122 and correlation information 123 of the second embodiment are the same as those of the first embodiment.

[0093] FIG. 13 is a diagram illustrating an example of the database 120 according to the second embodiment.

[0094] The data structure of the database 120 in the second embodiment is the same as that in the first embodiment. However, in the second embodiment, a new attribute value "D" is defined. The attribute value "D" indicates that it is not necessary to determine whether or not a matter related to a tag applies.

[0095] FIG. 14 is a diagram illustrating an example of the generation policy information 121 according to the second embodiment.

[0096] In the second embodiment, an entry is added that is a combination of the attribute value "D" of the tag of the data and the attribute value of the tag acquired from the search request.

[0097] In Example 2, the attribute value "D" is set in cases that also include tag-related matters. This eliminates the need to check the data content of data whose tag attribute value is "1." Therefore, in Example 2, the generation policies 404 for relationships "B" and "C" are "tag check" and "tag change." Furthermore, the generation policy for questions whose attribute value (data) 402 is "D" is "not required (D)."

[0098] The search process, tag analysis process, and question generation process of the second embodiment are the same as those of the first embodiment. Here, the process of the second embodiment will be described using a specific example.

[0099] In S101, when attribute values ​​such as those shown in Fig. 9 are acquired, the tag analysis process generates analysis information 1000 such as that shown in Fig. 15. Furthermore, the tag analysis process generates information such as that shown in Fig. 16.

[0100] In the process of S301 of the question generation process, tags for which the total number of "unnecessary (D)" generation policies is greater than a predetermined threshold may be excluded from selection. In the second embodiment, since confirmation of data content is not required, one of the indices (indicator 2), (indicator 3), or (indicator 4) is calculated in the question generation process.

[0101] Conventionally, either "applicable" or "inapplicable" was set as the attribute value, but in the second embodiment, by defining an attribute value that indicates that it is not necessary to judge whether the tag-related matter is applicable or not, it is possible to reduce the number of tags for which the attribute value needs to be checked. This is expected to speed up search processing and improve search accuracy.

[0102] The present invention is not limited to the above-described embodiments, but includes various modifications. For example, the above-described embodiments are provided to explain the present invention in detail, and the present invention is not necessarily limited to those including all of the described configurations. Furthermore, some of the configurations of each embodiment can be added to, deleted from, or replaced with other configurations.

[0103] Furthermore, the above-described configurations, functions, processing units, processing means, etc. may be partially or entirely implemented in hardware, for example, by designing them as integrated circuits. The present invention can also be realized by software program code that implements the functions of the embodiments. In this case, a storage medium on which the program code is recorded is provided to a computer, and a processor included in the computer reads the program code stored in the storage medium. In this case, the program code itself read from the storage medium implements the functions of the above-described embodiments, and the program code itself and the storage medium on which it is stored constitute the present invention. Examples of storage media for providing such program code include flexible disks, CD-ROMs, DVD-ROMs, hard disks, solid-state drives (SSDs), optical disks, magneto-optical disks, CD-Rs, magnetic tapes, non-volatile memory cards, and ROMs.

[0104] Furthermore, the program code that realizes the functions described in this embodiment can be implemented in a wide range of program or script languages, such as assembler, C / C++, perl, Shell, PHP, Python, and Java (registered trademark).

[0105] Furthermore, the program code of the software that realizes the functions of the embodiments may be distributed via a network and stored in a storage means such as a computer's hard disk or memory, or in a storage medium such as a CD-RW or CD-R, and the processor of the computer may read and execute the program code stored in the storage means or storage medium.

[0106] In the above-described embodiment, the control lines and information lines are those that are considered necessary for the explanation, and not all control lines and information lines are necessarily shown in the product. All components may be interconnected. [Explanation of symbols]

[0107] 100 Search System 101 terminals 110 Data Management Department 111 Search Department 120 databases 121 Generation policy information 122 Category Information 123 Correlation Information 200 calculator 201 processor 202 Network Interface 203 Main storage 204 Secondary storage device 205 Bus 1000 analysis information

Claims

1. 1. A computer system comprising: Connect to and access multiple databases that store data. The data is assigned a plurality of tags to be used for searching; The tag is assigned either a first attribute value representing that the tag corresponds to a matter related to the tag or a second attribute value representing that the tag does not correspond to a matter related to the tag, The computer system When a search request including information indicating whether a matter related to at least one of the tags is relevant is received, the attribute values ​​of the plurality of tags are obtained from the search request; Analyzing the relationship between the attribute values ​​of the plurality of tags of the data and the attribute values ​​of the plurality of acquired tags; referencing the database based on the results of the analysis to retrieve the data to be output as a response to the search request; generating a question for confirming the attribute value of the tag based on the result of the analysis; A computer system that outputs at least one of the retrieved data and the question sentence.

2. 2. The computer system of claim 1, The relationships include a first relationship indicating that it is necessary to check the attribute value of the tag, a second relationship indicating that it is necessary to check the attribute value of another tag that is related to the tag, and a third relationship indicating that it is not necessary to check the attribute value of the tag, The computer system In the analysis, a first process is performed to identify and record the relationship between each of the plurality of tags for each of the plurality of data; In generating the question, executing a second process of calculating, for each of the plurality of tags based on a result of the first process, the number of pieces of data for which at least one of the first relationship and the second relationship is established, as a first count value; executing a third process of calculating a sum of the first count values ​​for each of the plurality of tags as a first index based on a result of the second process; performing a fourth process of selecting the tag to be used to generate the question based on the first index; A computer system characterized by executing a fifth process that generates at least one of a question for confirming the attribute value of the selected tag and a question for confirming the attribute value of another tag that is related to the selected tag.

3. 3. The computer system according to claim 2, calculating, for each of the plurality of data, the number of tags for which the relationship is established, as a second count value, based on a result of the first processing; A computer system characterized in that the data to be analyzed is selected based on the second count value and the attribute values ​​of the multiple tags obtained from answers to the question.

4. 3. The computer system according to claim 2, The tag is assigned one of the first attribute value, the second attribute value, and a third attribute value indicating that it is not necessary to determine whether a matter related to the tag applies; the relationships include the first relationship, the second relationship, the third relationship, and a fourth relationship representing that the attribute value of the tag of the data is the third attribute value; The computer system In the second process, the number of pieces of data for which the fourth relationship is established is calculated as a third count value for each of the plurality of tags based on a result of the first process; In the third process, a sum of the third count values ​​is calculated as a second index for each of the plurality of tags based on a result of the second process; The fourth process is characterized in that the tags to be excluded from selection are selected based on the second indicator, and the tags to be used to generate the question sentence are selected from the excluded tags based on the first indicator.

5. 3. The computer system according to claim 2, A computer system that selects the question sentence to be output based on the first index.

6. A data search method executed by a computer system, comprising: the computer system is connected to a database for storing a plurality of data so as to be accessible; The data is assigned a plurality of tags to be used for searching; The tag is assigned either a first attribute value representing that the tag corresponds to a matter related to the tag or a second attribute value representing that the tag does not correspond to a matter related to the tag, The data search method is a first step in which the computer system acquires the attribute values ​​of the plurality of tags from the search request when the computer system receives a search request including information indicating whether a matter related to the tag is relevant for at least one of the tags; a second step in which the computer system analyzes a relationship between the attribute values ​​of the plurality of tags of the data and the attribute values ​​of the plurality of acquired tags; a third step in which the computer system refers to the database based on the results of the analysis and searches for the data to be output as a response to the search request; a fourth step in which the computer system generates a query for confirming the attribute value of the tag based on the result of the analysis; a fifth step in which the computer system outputs at least one of the searched data and the query.

7. 7. A data search method according to claim 6, The relationships include a first relationship indicating that it is necessary to check the attribute value of the tag, a second relationship indicating that it is necessary to check the attribute value of another tag that is related to the tag, and a third relationship indicating that it is not necessary to check the attribute value of the tag, the second step includes a step of executing a first process by the computer system to identify and record the relationship between each of the plurality of tags for each of the plurality of data; The fourth step includes: a step of executing a second process by the computer system to calculate, based on a result of the first process, the number of pieces of data for which at least one of the first relationship and the second relationship is established for each of the plurality of tags, as a first count value; a step of executing a third process by the computer system to calculate, as a first index, a sum of the first count values ​​for each of the plurality of tags based on a result of the second process; a step in which the computer system executes a fourth process of selecting the tag to be used to generate the question sentence based on the first index; A data search method characterized in that the computer system includes a step of executing a fifth process to generate at least one of a question sentence for confirming the attribute value of the selected tag and a question sentence for confirming the attribute value of another tag that is related to the selected tag.

8. 8. A data search method according to claim 7, The fourth step includes: a step in which the computer system calculates, for each of the plurality of data, the number of tags for which the relationship is established, as a second count value, based on a result of the first processing; A data search method characterized by including a step in which the computer system selects the data to be analyzed based on the second count value and the attribute values ​​of the multiple tags obtained from the answers to the question.

9. 8. A data search method according to claim 7, The tag is assigned one of the first attribute value, the second attribute value, and a third attribute value indicating that it is not necessary to determine whether a matter related to the tag applies; the relationships include the first relationship, the second relationship, the third relationship, and a fourth relationship representing that the attribute value of the tag of the data is the third attribute value; The computer system the second processing includes a step of the computer system calculating, based on a result of the first processing, the number of pieces of data for which the fourth relationship is established for each of the plurality of tags as a third count value; the third processing includes a step of the computer system calculating, based on a result of the second processing, a sum of the third count values ​​for each of the plurality of tags as a second index; The fourth processing is a data search method characterized in that the computer system includes a step of selecting the tags to be excluded from selection based on the second indicator, and selecting the tags to be used to generate the question sentence from among the excluded tags based on the first indicator.

10. 8. A data search method according to claim 7, The data search method is characterized in that the fourth step includes a step in which the computer system selects the query sentence to be output based on the first index.

Images