Document search / generation system and program

Abstract

To provide a document search / generation system and program capable of searching for document data desired by a user from among multiple pieces of document data stored in a specific storage area, and then easily generating document data including content responding to user's needs based on the retrieved document data, thereby effectively utilizing the stored document data.SOLUTION: Provided is a document search / generation system and program, in which a document search / generation server 1 causes an index storage unit 22 to store index data from original document data of a document data storage unit 21, interactively searches for segmented text data in the index data stored in the index storage unit 22 based on search content from a user terminal 3, generates a document according to a generation instruction from a user based on the segmented text data that have been extracted by searching, and outputs generated document data to the user terminal 3.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to a document search and generation system and program that interactively searches for target document data from multiple document data sets and generates document data with the answer content desired by the user from the searched document data. [Background technology]

[0002] [Prior Art] In conventional document search systems, target document data is searched for using keywords. There were also some that used generative AI (artificial intelligence) to obtain answers in a dialogue (chat) format.

[0003] [Related Technology] Related prior art includes Japanese Patent Application Publication No. 62-236071, entitled "Document Creation Processing Apparatus" (Patent Document 1), Japanese Patent Application Publication No. 04-369078, entitled "In-house Patent Abstract Creation System" (Patent Document 2), Japanese Patent Application Publication No. 2008-077459, entitled "Interactive Multiple Document Summarization Apparatus" (Patent Document 3), Japanese Patent Application Publication No. 2018-173681, entitled "Search Result Summarization Apparatus, Program, and Method" (Patent Document 4), and Japanese Patent Application Publication No. 2023-120130, entitled "Conversational AI Platform Using Extracted Question Answering" (Patent Document 5).

[0004] Patent Document 1 discloses a document creation device that displays variably displayed items of documents for each process on a common input screen and edits and creates documents.

[0005] Patent Document 2 discloses a system that retrieves bibliographic data from a patent management database (DB), retrieves invention data contents from an application DB, and creates an abstract from the two retrieved data.

[0006] Patent Document 3 discloses an apparatus that interactively generates a plurality of summary documents from a plurality of documents, and regenerates an abstract based on a portion of interest selected by an operator from the plurality of summary documents.

[0007] Patent Document 4 discloses a device that extracts and summarizes the content of each content included in an internet search result, and generates and outputs a document showing the summary information.

[0008] Patent Document 5 shows that a conversational AI platform uses structured data and unstructured data to respond to queries from users, and searches and responds in the unstructured data. [Prior art documents] [Patent documents]

[0009] [Patent Document 1] Japanese Patent Application Publication No. 62-236071 [Patent Document 2] Japanese Patent Application Publication No. 04-369078 [Patent Document 3] Japanese Patent Application Laid-Open No. 2008-077459 [Patent Document 4] Japanese Patent Application Publication No. 2018-173681 [Patent Document 5] Japanese Patent Publication No. 2023-120130 Summary of the Invention [Problem to be solved by the invention]

[0010] However, the above technology does not search for the desired document data from among the document data stored in a specific, reliable storage area and generate document data based on that data that will provide the answer the user desires, and therefore has the problem of not being able to easily obtain the document data the user desires from highly reliable document data.

[0011] Patent Documents 1 to 5 do not describe a configuration for interactively searching for target document data from a plurality of document data and generating document data with answer contents desired by the user from the searched document data.

[0012] The present invention has been made in consideration of the above-mentioned circumstances, and aims to provide a document search and generation system and program that can search for document data desired by a user from multiple document data stored in a specific storage area, and then easily generate document data that contains the answer the user is looking for based on the searched document data, thereby making effective use of the stored document data. [Means for solving the problem]

[0013] The present invention, which solves the problems of the above-mentioned conventional examples, is a document search and generation system having a document search and generation server that searches for documents in response to a user request and generates an answer, wherein the document search and generation server has a document data storage unit that stores document data to be searched in a specific folder, and an index storage unit that stores index data that associates, for the document data in the folder, the file names of the document data with segmented text data into which text data extracted from the document data has been segmented into multiple pieces, and when search content and generation instructions are input interactively from a user's user terminal, the segmented text data is searched for in accordance with the search content, a document of the answer content is generated based on the searched segmented text data in accordance with the generation instructions, and the generated document data is displayed and output on the user terminal.

[0014] The present invention is characterized in that, in the above document search and generation system, when the document search and generation server generates a document, it displays the file name of the document data corresponding to the divided text data on the user terminal, and when the file name is selected on the user terminal, it reads the document data of that file name from the document data storage unit and displays it on the user terminal.

[0015] The present invention is characterized in that in the document search and creation system, the document search and creation server assigns a page number for reference use to the file name of the document data displayed on the user terminal.

[0016] The present invention is characterized in that in the document search and creation system, the document search and creation server assigns a relevance score to the file name of the document data displayed on the user terminal.

[0017] The present invention is characterized in that in the above document search and generation system, the document search and generation server has a plurality of folders in the document data storage unit, a plurality of folders in the index storage unit corresponding to the plurality of folders, access rights to the plurality of folders in the document data storage unit are set for each user, and the document search and generation server accepts requests from user terminals of users who have access rights to the plurality of folders.

[0018] The present invention is characterized in that in the above document search and generation system, the document search and generation server assigns an access folder to the file name in the index storage unit and associates an access list that allows access from the user terminal with the access folder.

[0019] The present invention is a processing program that operates on a document search and generation server that searches for documents and generates answers in response to user requests, and is characterized by causing the document search and generation server to store the document data to be searched in a specific folder in a document data storage unit, extract text data from the document data in the folder, and store index data in the index storage unit that associates the file names of the document data with the divided text data divided into multiple pieces, and when search content and generation instructions are input interactively from a user's user terminal, search for the divided text data in accordance with the search content, generate a document of the answer content based on the searched divided text data in accordance with the generation instructions, and display and output the generated document data on the user terminal.

[0020] The present invention is characterized in that, in the above program, the document search and generation server is configured to display the file name of the document data corresponding to the divided text data on the user terminal when a document is generated, and when the file name is selected on the user terminal, read the document data of that file name from the document data storage unit and display it on the user terminal.

[0021] The present invention is characterized in that the program causes the document search and creation server to function so as to add a page number for reference use to the file name of the document data displayed on the user terminal.

[0022] The present invention is characterized in that the program causes the document search and creation server to function to assign a relevance score to the file name of the document data displayed on the user terminal.

[0023] The present invention is characterized in that, in the above program, the document search and generation server is configured to provide multiple folders in the document data storage unit, provide multiple folders in the index storage unit corresponding to the multiple folders, set access rights to the multiple folders in the document data storage unit for each user, and function to accept requests from user terminals of users who have access rights to the multiple folders.

[0024] The present invention is characterized in that, in the above program, the document search and generation server is made to function in the index storage unit so as to assign an access folder to a file name and associate an access list that allows access from a user terminal with the access folder. [Effects of the Invention]

[0025] According to the present invention, the document search and generation server is equipped with a document data storage unit that stores document data to be searched in a specific folder, and an index storage unit that stores index data that associates the file names of the document data in the folder with split text data into multiple pieces of text data extracted from the document data.When search content and generation instructions are input interactively from the user's user terminal, the split text data is searched according to the search content, a document of answer content is generated based on the searched split text data in accordance with the generation instructions, and the generated document data is displayed and output on the user terminal.This document search and generation system allows the user to easily generate a new answer document as requested from reliable document data in a specific folder in an interactive manner, and has the effect of being able to effectively utilize the document data stored in the specific folder. [Brief explanation of the drawings]

[0026] [Figure 1] FIG. 1 is a schematic diagram of the system. [Figure 2] FIG. 1 is a schematic diagram of the processing of the present system. [Figure 3] FIG. 10 is a schematic diagram of an example of a display on a user terminal. [Figure 4] FIG. 2 is a schematic diagram showing user access rights. [Figure 5] FIG. 10 is a schematic diagram of an example of page number display of a reference document. [Figure 6] FIG. 10 is a schematic diagram of an example of displaying a document number in a response document. [Figure 7] FIG. 10 is a schematic diagram of an example of displaying relevance scores. [Figure 8] FIG. 1 is a schematic diagram illustrating an access structure. [Figure 9] FIG. 10 is a schematic diagram illustrating another access structure. DETAILED DESCRIPTION OF THE INVENTION

[0027] An embodiment of the present invention will be described with reference to the drawings. [Outline of the embodiment] The document search and generation system (this system) according to an embodiment of the present invention stores multiple document data in a specific memory area in advance, and the document search and generation server interactively searches the specific memory area for the document data that the user is searching for, generates a document with the answer content requested by the user based on the retrieved document data, and provides it to the user terminal.The user can easily interactively generate a new answer document that he or she requests from reliable document data in the specific memory area, and effectively utilize the document data stored in the specific memory area.

[0028] In addition, in this system, the document data of the newly generated answer content is displayed on the user terminal, and the name of the referenced document data (file name) is also displayed. When the document data name is selected, the referenced document data is read from a specific memory area and displayed, making it easy to check the document data that referenced the document of the generated answer content.

[0029] [This system: Figure 1] This system will be described with reference to Figure 1. Figure 1 is a schematic diagram of the system. As shown in FIG. 1, this system comprises a document search and creation server 1, a user terminal 3, and a network 4. The document search and creation server 1 and the user terminal 3 are connected via a network 4 .

[0030] Each part of this system will now be described in detail. [Document search and creation server 1] The document search and generation server 1 stores multiple document data (original document data) that are candidates for search targets in the document data storage unit 21, generates index data based on the original document data, and stores it in the index storage unit 22. The index data is composed of the file name of the original document data and divided text data obtained by extracting text data from the original document data and dividing it into a plurality of units.

[0031] When the search content specifying the search target and the instructions for document generation are input from the user terminal 3 in an interactive (chat) format, the document search and generation server 1 searches the index data stored in the index storage unit 22 to extract the corresponding divided text data, generates a document of the answer content using a generation AI based on the divided text data extracted in accordance with the instructions for document generation from the user, and displays the generated document data (generated document data) on the user terminal 3.

[0032] If there are multiple pieces of divided text data to be extracted depending on the search content, the answer content document is generated by referring to the multiple pieces of divided text data. The document data search and extraction process and document generation process will be described in detail later.

[0033] The document search and creation server 1 also includes a control unit 11, a storage unit 12, and an interface unit 13, and the interface unit 13 is connected to the network 4, a document data storage unit 21, and an index storage unit 22. The document data storage unit and the index storage unit may be provided within the storage unit 12.

[0034] The storage unit 12 stores processing programs for realizing the characteristic processing of this system, other necessary parameters, and the like. The control unit 11 reads and executes a processing program stored in the storage unit 12 to realize processing functions described below.

[0035] [Document data storage unit 21] The document data storage unit 21 stores, in folder units, a plurality of original document data items that are referenced and used to generate documents of response contents. The folder has access rights set for each user, which will be described later.

[0036] [Index storage unit 22] The index storage unit 22 has folders corresponding to the folders of the document data storage unit 21, and stores in these folders as index data the document names (file names) of the original document data stored in the document data storage unit 21 and text data (divided text data) obtained by dividing the text of the document data into multiple parts.

[0037] The divided text data is extracted from the original document data and separated into specific coherent items. The page number of the original document data is assigned as an attribute of the divided text data. The use of the page number will be explained in the application example below.

[0038] Further, the folders in the document data storage unit 21 and the folders in the index storage unit 22 are set with access rights that allow users to access them. User authority is set for each folder in the document data storage unit 21 in correspondence with the user ID, and is stored in the document search and creation server 1. If the user authority permits access to the folder in the document data storage unit 21 to be searched, the divided text data in the corresponding folder in the index storage unit 22 can be searched.

[0039] In the above example, the user's access authority is set on a folder-by-folder basis, but the user's access authority may also be set on an original document data basis or on an index data basis. In addition, in the above example, the folders in the document data storage unit 21 and the folders in the index storage unit 22 are associated with each other to set the user's access rights, but this association is not necessary. For example, the range of accessibility to the folders in the index storage unit 22 may be widened to improve the accuracy of generating generated document data.

[0040] The document search process is performed on a segmented text data basis. Searching each segmented text data rather than the entire original document data makes it easier to quickly find documents related to the search content and to calculate the degree of relevance (correlation).

[0041] In this system, divided text data of index data is searched, but the text data portion of the original document data in the document data storage unit 21 may be searched directly.

[0042] [User terminal 3] The user terminal 3 is a processing device of a computer of a user who is permitted to use the services of this system, and is equipped with a display unit and an input unit, and is assumed to be a personal computer (PC), a tablet, or a smartphone.

[0043] The user accesses the document search and creation server 1 from the user terminal 3 using the user ID and password, and inputs the search content they wish to search and the creation instructions they wish to create in chat format. The created document data of the answer content generated based on the search results is then displayed on the display unit of the user terminal 3. The search content in chat format is, for example, "search for XX," and the generation instruction is, for example, "do XX based on the searched information."

[0044] The display unit of the user terminal 3 displays the file name of the generated document (answer document) and the original document data used for generation, and when the file name is selected, the corresponding original document data is read from the document data storage unit 21 and displayed for reference.

[0045] [Network 4] Network 4 is assumed to be a LAN (Local Area Network), but it may also be the Internet where users log in with an ID and password. This system is intended for use in a closed environment within a specific company or corporate group. It is also suitable for use within public institutions and other organizations, not just specific corporations.

[0046] Furthermore, folders in the document data storage unit 21 and folders in the index storage unit 22 may be provided for each department in a company or the like, so that they can be used on a departmental basis. Furthermore, a user may be allowed to use multiple folders in the document data storage unit 21 and multiple folders in the index storage unit 22. For example, folders in the document data storage unit 21 and the index storage unit 22 may be prepared separately for each department and for all employees, and the user may use both folders.

[0047] [Processing of this system: Figure 2] Next, the processing of this system will be described with reference to Fig. 2. Fig. 2 is a schematic diagram of the processing of this system. As a premise, the document search and creation server 1 stores a plurality of original document data that are candidates for search targets in the document data storage unit 21, for each folder, and further generates index data from the original document data in the folder and stores it in the corresponding folder in the index storage unit 22 (S0). The index data consists of a file name and a plurality of divided text data.

[0048] The user terminal 3 then accesses the document search and creation server 1 via the network 4 and inputs the search content and document creation instructions (S1), which causes the document search and creation server 1 to execute a search process (S2). The search process searches for divided text data in the index data in the index storage unit 22, calculates (scoring) the relevance (correlation) of the search content, and extracts divided text data with a score higher than a specific reference value (S3).

[0049] The document search and generation server 1 inputs the extracted divided text data, and generates generated document data that will serve as the answer content based on the divided text data using a generation AI in accordance with generation instructions from the user (S4), and outputs the generated document data to the user terminal 3 (S5).

[0050] Here, the document search and creation server 1 weights the segmented text data by the search score, and generates documents by increasing the usage rate of segmented text data with high scores and decreasing the usage rate of segmented text data with low scores. The user terminal 3 inputs the generated document data from the document search and creation server 1 and displays it on the display screen.

[0051] The document search and generation server 1 displays the file name searched for and used to generate the document in the generated document data output to the user terminal 3, and links the file name to the original document data in the document data storage unit 21. When the user selects the file name displayed in the generated document data by clicking or the like, the original document data is displayed on the display unit of the user terminal 3 so that it can be referenced.

[0052] [Example of display on user terminal 3: Figure 3] A display example displayed on the display unit of the user terminal 3 will be described with reference to Fig. 3. Fig. 3 is a schematic diagram of a display example on the user terminal. As shown in FIG. 3, a prompt such as "Search for XX and then do △△ based on the searched information" is input from the user terminal 3 as a question.

[0053] Then, the document search and generation server 1 performs a search according to the search content "Search for XX" and the generation instruction "Do △△ based on the searched information", generates generated document data of the answer content according to the generation instruction, displays the generated document data, and below it displays the file name (document name) of the original document data that was the basis for the generated document data (original document data used by reference). The file name is assigned a document number in order. The generated document data (answer document) may be a list of key points or a collection of sentences.

[0054] Then, when the file name of the original document data displayed on the display unit of the user terminal 3 is clicked to select it, the corresponding original document data is read from the document data storage unit 21 and displayed on the display unit of the user terminal 3. This makes it easy to check the original document data used to generate the generated document data.

[0055] [Access Permissions: Figure 4] In addition, the document search and creation server 1 sets access rights to original document data for each user on a folder-by-folder basis in the document data storage unit 21, and manages, for example, the association between user IDs and access rights to folders. Therefore, depending on the user, folders that can be searched for index data in the index storage unit 22 are limited, and the content of the generated document may also differ because the original document data that can be used differs depending on the authority.

[0056] The access rights to folders in the document data storage unit 21 will be described with reference to Fig. 4. Fig. 4 is a schematic diagram showing the access rights of users. The document search and creation server 1 stores access rights indicating whether or not a user can access a folder in the document data storage unit 21 shown in FIG. 4 in the form of a table or the like in the storage unit 12.

[0057] In Figure 4, the settings are such that user ID "A" can access folders A and C in the document data storage unit 21 but cannot access folder B, user ID "B" can access folders A and BC, and user ID "C" can access folder A but cannot access folders B and C. Here, the user IDs "A", "B", and "C" indicate the login IDs of users A, B, and C.

[0058] The contents of the user's access rights in Figure 4 also apply to folders in the index storage unit 22 that correspond to folders in the document data storage unit 21, and when searching for index data, access to folders in the index storage unit 22 is similarly restricted. That is, the user can search for index data in a folder in the index storage unit 22 that corresponds to a folder in the document data storage unit 21 that the user is permitted to access.

[0059] Therefore, when a user logs in to the service of this system from a user terminal 3, the document search and generation server 1 checks whether there is a folder in the document data storage unit 21 that can be accessed by that user ID, and if there is, it accepts the request from the user, searches the folder in the index storage unit 22 that corresponds to that folder according to the search content, and generates generated document data of the answer content.

[0060] [Application example] Next, an application example of this system will be described. [Keyword highlighting] When the document search and creation server 1 displays the created document data on the display unit of the user terminal 3, the document search and creation server 1 may highlight keywords extracted in relation to the search from the searched divided text data. Additionally, keywords entered in the question may also be highlighted. It is desirable to display the question keywords and the extracted keywords in different highlight colors.

[0061] [Specify reference document data] When a question is input from the user terminal 3, the document search and generation server 1 allows a specific document name to be specified as a search condition, and when a document name is specified, the document search and generation server 1 searches for index data corresponding to the document name and displays the answer content as generated document data, displays the specified document name in the column below the generated document data, and allows the document content to be viewed by selecting the document name.

[0062] [Page numbers of reference documents: Figure 5] As shown in Figure 5, the document search and generation server 1 may display the page number used in the divided text data in the document name displayed in the column below the generated document data displayed on the display unit of the user terminal 3, and when the page number is selected by clicking or the like, the corresponding page of the original document data for the document name will be displayed. FIG. 5 is a schematic diagram of an example of page number display of a reference document.

[0063] Since page numbers are assigned as attribute data to the divided text data in the index storage unit 22, the document search and creation server 1 can display the page numbers next to the document names.

[0064] When a document name and page number are specified from the user terminal 3, the document search and creation server 1 instructs the document data storage unit 21 to display the relevant part by instructing the document name (file name) and page number, thereby enabling easy reference to parts related to the created document data.

[0065] [Display the document number in the response document: Figure 6] As shown in Fig. 6, when the answer document is itemized in the generated document data (answer document) displayed on the display unit of the user terminal 3, the document search and creation server 1 may display the referenced document number to the right of the itemized sentences. This makes it easy to recognize that the itemized sentences in the generated document data are based on the document number displayed next to them. FIG. 6 is a schematic diagram of an example of how a document number is displayed in a response document.

[0066] [Displaying the relevance score: Figure 7] 7, the document search and creation server 1 may display a score of relevance (degree of relevance) next to the document name displayed in the column below the created document data displayed on the display unit of the user terminal 3. This allows the user to easily recognize the degree of relevance of the reference document. FIG. 7 is a schematic diagram of an example of displaying the relevance score. Although the relevance (correlation to the question) is displayed numerically, the degree to which the information was referenced (used) to generate the answer may be displayed numerically instead of the relevance.

[0067] [Access structure: Figure 8] The access structure in this system will be described with reference to Fig. 8. Fig. 8 is a schematic diagram showing the access structure. As mentioned above, the index storage unit 22 has folders corresponding to the folders in the document data storage unit 21, and a list (access list) of user IDs or email addresses of users with access rights is set for each piece of divided text data. Hereinafter, the access list will be described as email addresses.

[0068] Specifically, as shown in FIG. 8(a), the index storage unit 22 stores divided text data and an access list in association with a file name. The access list is a list of email addresses that are permitted to access the file. In the structure of FIG. 8(a), an access list is set for each piece of divided text data, which can result in a redundant data structure.

[0069] Therefore, as shown in FIG. 8(b), a folder name for access and divided text data are set in association with the file name, and the access list is deleted and stored in the index storage unit 22. Specifically, in Figure 8(b), the file name "File (1)" is associated with the access folder name "Folder (1)," and the divided text data "Text (1-01)," "Text (1-02)," and "Text (1-03)" are associated with "File (1)," the file name "File (2)" is associated with the folder name "Folder (2)," and the divided text data "Text (2-01)" and "Text (2-02)" are associated with "File (2)."

[0070] Then, as shown in FIG. 8(c), the access list is stored in the index storage unit 22 as an access table in association with the folder name. Specifically, in FIG. 8(c), "Access List (1)" is associated with the access folder name "Folder (1)," and "Access List (2)" is associated with "Folder (2)."

[0071] In this system, when a request for document search and generation service is received from user terminal 3, the system references the email address in the access list of the access table, accesses and searches the divided text data of the file name corresponding to the access folder name set for the user's email address, and generates the document.

[0072] By adopting the data structure shown in FIG. 8(b)(c), an access list is stored for each folder name used for access, which simplifies the data structure. 8(b) and 8(c), one folder name is associated with the file name of the file corresponding to the text data before division, but one folder name may be assigned to multiple file names. Another access structure will be explained below.

[0073] [Another access structure: Figure 9] Another access structure will be described with reference to Figure 9. Figure 9 is a schematic diagram showing another access structure. Another access structure may be as shown in Figure 9(a), in which the folder name for access, "Folder (1)," is associated with not only the file name "File (1)" but also "File (3)," and the divided text data "Text (3-01)" and "Text (3-02)" are associated with "File (3)."

[0074] Conversely, one file name may be associated with multiple access folders. Specifically, as shown in Figure 9(b), the file name "File (4)" and the divided text data "Text (4-01)" may be assigned the folder name "Folder (4)" for access purposes, and the file name "File (4)" and the divided text data "Text (4-02)," "Text (4-03)," and "Text (4-04)" may be assigned the folder name "Folder (5)" for access purposes, and different access lists may be associated with each of them.

[0075] In the example of Figure 9(b), for example, if the divided text data "Text (4-01)" is the summary part of the original document data, and the divided text data "Text (4-02)" to "Text (4-04)" are the detailed content parts of the original document data, the access authority for the summary part and the detailed content part is changed. The access table will be the same as Figure 8(c). In other words, the degree of freedom in assigning folder names for access to file names allows for flexible adaptation to the user's environment.

[0076] [Effects of the embodiment] According to this system, a plurality of original document data are stored in the document data storage unit 21 in advance, index data consisting of file names and divided text data is created and stored in the index storage unit 22 from the original document data, the document search and generation server 1 interactively searches for the divided text data based on the search contents from the user terminal 3, generates a document based on the searched and extracted divided text data in accordance with the generation instructions from the user, and outputs the generated document data to the user terminal 3. This allows the user to easily generate the desired generated document data from the reliable original document data in the document data storage unit 21 in an interactive manner, and has the effect of making effective use of the original document data stored in the document data storage unit 21. [Industrial Applicability]

[0077] The present invention is suitable for a document search and generation system and program that can search for document data desired by a user from multiple document data stored in a specific storage area, and then easily generate document data containing the answer desired by the user based on the searched document data, thereby making effective use of the stored document data. [Explanation of symbols]

[0078] 1... document search and creation server, 3... user terminal, 4... network, 11... control unit, 12... storage unit, 13... interface unit, 21... document data storage unit, 22... index storage unit,

Claims

1. A document search and generation system having a document search and generation server that searches for documents and generates answers in response to user requests, The document search and creation server, a document data storage unit that stores document data to be searched in a specific folder; an index storage unit that stores index data that associates file names of the document data in the folder with divided text data obtained by dividing text data extracted from the document data into a plurality of pieces, When search content and generation instructions are input interactively from the user terminal of the user, the divided text data is searched in accordance with the search content, a document of the answer content is generated based on the searched divided text data in accordance with the generation instructions, and the generated document data is displayed on the user terminal.

2. The document search and generation system described in claim 1, characterized in that when the document search and generation server generates a document, it displays the file name of the document data corresponding to the divided text data on the user terminal, and when the file name is selected on the user terminal, it reads the document data of the file name from the document data storage unit and displays it on the user terminal.

3. 3. The document search and generation system according to claim 2, wherein said document search and generation server adds a page number for reference use to a file name of the document data displayed on said user terminal.

4. 3. The document search and generation system according to claim 2, wherein the document search and generation server assigns a relevance score to the file name of the document data displayed on the user terminal.

5. The document search and generation system according to claim 1 or 2, characterized in that the document search and generation server has a plurality of folders in the document data storage unit, a plurality of folders in the index storage unit corresponding to the plurality of folders, access rights to the plurality of folders in the document data storage unit are set for each user, and the document search and generation server accepts requests from user terminals of users who have access rights to the plurality of folders.

6. The document search and generation system according to claim 1 or 2, characterized in that the document search and generation server assigns an access folder to the file name in the index storage unit and associates an access list that allows access from the user terminal with the access folder.

7. A processing program that runs on a document search and generation server that searches for documents and generates answers in response to user requests, The document search and creation server, storing document data to be searched in a specific folder in a document data storage unit; extracting text data from the document data in the folder, and storing index data that associates the file names of the document data with the divided text data into a plurality of pieces in an index storage unit; When search content and generation instructions are input interactively from the user terminal of the user, the program searches the divided text data according to the search content, generates a document of the answer content based on the searched divided text data according to the generation instructions, and displays the generated document data on the user terminal.

8. The program of claim 7, characterized in that the document search and generation server is configured to, when a document is generated, display the file name of the document data corresponding to the divided text data on the user terminal, and when the file name is selected on the user terminal, read the document data of the file name from the document data storage unit and display it on the user terminal.

9. 9. The program according to claim 8, wherein the document search and creation server is caused to function to add a page number for reference use to a file name of document data displayed on the user terminal.

10. 9. The program according to claim 8, wherein the document search and creation server is caused to function to assign a relevance score to a file name of document data displayed on the user terminal.

11. The program of claim 7 or 8, characterized in that the document search and generation server is configured to provide a plurality of folders in the document data storage unit, provide a plurality of folders in the index storage unit corresponding to the plurality of folders, set access rights to the plurality of folders in the document data storage unit for each user, and accept requests from user terminals of users who have access rights to the plurality of folders.

12. The program according to claim 7 or 8, characterized in that the document search and generation server is caused to function in the index storage unit to assign an access folder to the file name and associate an access list that allows access from a user terminal with the access folder.

Images