Intelligent interaction method and system of document

Abstract

The invention discloses an intelligent interaction method and system for documents. The method comprises the steps of receiving documents uploaded by a user, and generating a document list; receiving a reading request of a user for a target document in the document list, and generating a session type reading interaction interface which comprises a tool area, a document outline area, a current document reading area and a session window area; generating and displaying a document abstract of the target document in the session window area through a document assistant, wherein the document assistant comprises a ChatGPT natural language processing model; performing real-time monitoring on content selection of the target document by a user in the current document reading area, and obtaining questions of the user; and generating an answer of the question through the document assistant and displaying the answer in the session window area. The invention provides a document reading scheme capable of realizing intelligent interaction, so that the reading experience of the user is improved, and the reading requirement of the user can be better met.

Description

Technical Field

[0001] The present invention relates to the technical field of document processing, and in particular to an intelligent document interaction method and system. Background Art

[0002] With the rapid development of Internet technology, online document management and sharing have become increasingly popular. Taking PDF documents as an example, PDF has become one of the most commonly used document formats due to its cross-platform compatibility and format consistency. Users not only want to view PDF documents online on multiple terminals, but also want customized reading based on user needs, using AI to analyze document content to improve document processing efficiency and reading experience. Traditional PDF reading tools have an outdated interface and only provide basic reading functions, which is not convenient for expanding personalized function options.

[0003] Therefore, the prior art still needs to be improved and developed. Summary of the invention

[0004] The main purpose of the present invention is to solve the technical problems mentioned in the above background technology existing in the existing document reading tools.

[0005] A first aspect of the present invention provides a document intelligent interaction method, comprising:

[0006] Receive documents uploaded by users and generate a document list;

[0007] Receiving a user's reading request for a target document in the document list, and generating a conversational reading interaction interface, wherein the conversational reading interaction interface includes a tool area, a document outline area, a document current reading area, and a conversation window area;

[0008] Generate and display a document summary of the target document in the conversation window area by means of a document assistant, wherein the document assistant includes a ChatGPT natural language processing model;

[0009] Real-time monitoring of the user's selection of the content of the target document in the current reading area of ​​the document and obtaining the user's questions;

[0010] The answer to the question is generated by the document assistant and displayed in the conversation window area.

[0011] In an optional implementation of the first aspect of the present invention, the receiving a user's reading request for a target document in the document list and generating a conversational reading interaction interface includes:

[0012] Receiving a user's request to read a target document in the document list, and obtaining a document link of the target document and initialization data of the conversational reading interaction interface;

[0013] Generate a partition framework of the conversational reading interaction interface based on the initialization data;

[0014] generating a tool set in a tool area of ​​the partition framework;

[0015] Obtaining a directory of the target document based on the link;

[0016] Generate a hierarchically foldable and expandable outline tree in the document outline area of ​​the partition framework based on the directory;

[0017] The first page of the target document is displayed in the current reading area of ​​the document.

[0018] In an optional implementation of the first aspect of the present invention, the generating and displaying the document summary of the target document in the conversation window area by the document assistant includes:

[0019] Determine whether the page displayed in the current reading area of ​​the document is the homepage of the target document;

[0020] If the current page displayed in the current reading area of ​​the document is the homepage of the target document, the summary page of the target document is obtained through the document assistant, and the full text summary of the target document is displayed in the session window area;

[0021] If the current page displayed in the current reading area of ​​the document is not the homepage of the target document, the document assistant reads the content and semantically understands the current page displayed in the current reading area of ​​the document, obtains a page summary of the current page, and displays the page summary in the session window area.

[0022] In an optional implementation of the first aspect of the present invention, the real-time monitoring of the user's selection of the content of the target document in the current reading area of ​​the document and obtaining the user's question includes:

[0023] Get the cursor style of the current reading area of ​​the document;

[0024] Determine whether the cursor style is a text selection cursor;

[0025] When the cursor style is the text selection cursor, a highlight selection function is bound to the text selection cursor;

[0026] Performing real-time monitoring on the moving path of the text selection cursor;

[0027] When the text selection cursor is selecting a word, the word symbol that the moving path passes through is converted into a highlight display style;

[0028] Obtain a text set in the highlighted display style and generate a question input pop-up window;

[0029] The question input pop-up window is used to receive questions input by the user.

[0030] In an optional implementation manner of the first aspect of the present invention, generating the answer to the question by the document assistant and displaying it in the conversation window area includes:

[0031] Extracting keywords from the text set of the question by the document assistant;

[0032] Performing knowledge extension on the text set based on the keywords;

[0033] Acquire the semantics of the question, and obtain an answer to the question from the knowledge extension based on the semantics;

[0034] The text set of the question and the answer are bound and displayed in the conversation window area.

[0035] In an optional implementation of the first aspect of the present invention, after the text set of the question and the answer are displayed in the conversation window area in a binding manner, the method further includes:

[0036] Acquiring, by the document assistant, a number of related questions from the historical question records of the text set or the question association relationship library;

[0037] A number of the related questions are displayed below the answer, so that the user can click on them to obtain a more in-depth answer through the document assistant or other means.

[0038] In an optional implementation of the first aspect of the present invention, the document uploading method includes local document uploading and URL link uploading;

[0039] If the document uploading method is local document uploading, the receiving of the document uploaded by the user and the generating of the document list include:

[0040] Receive the document selected by the user based on clicking the upload local document option;

[0041] Conduct legal and compliance checks on the documents and verify user rights and interests;

[0042] After the legal compliance detection and the user rights verification are passed, the document is sliced ​​and uploaded to the OSS system for unified storage management, and a document list is generated;

[0043] If the document is uploaded in a URL link upload mode, the step of receiving the document uploaded by the user and generating a document list includes:

[0044] Receive URL links uploaded by users;

[0045] Conduct legal and compliance checks on the URL link and verify user rights;

[0046] After the legal compliance detection and the user rights verification are passed, the document is downloaded based on the URL link, and the document format and document size are verified;

[0047] After the document format and the document size are verified, the document will be sliced ​​and uploaded to the OSS system for unified storage management, and a document list will be generated.

[0048] A second aspect of the present invention provides a document intelligent interaction system, the intelligent interaction system comprising:

[0049] The document receiving module is used to receive documents uploaded by users and generate a document list;

[0050] A reading request receiving module, used to receive a user's reading request for a target document in the document list, and generate a conversational reading interaction interface, wherein the conversational reading interaction interface includes a tool area, a document outline area, a document current reading area, and a conversation window area;

[0051] A summary generation module, used to generate and display a document summary of the target document in the conversation window area through a document assistant, wherein the document assistant includes a ChatGPT natural language processing model;

[0052] A question acquisition module, used to monitor the user's selection of the content of the target document in the current reading area of ​​the document in real time and acquire the user's question;

[0053] The answer display module is used to generate the answer to the question through the document assistant and display it in the conversation window area. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 The figure is a flowchart of an exemplary method for intelligent interaction of documents according to the present invention.

[0055] Figure 2 The figure is an exemplary flow chart of the interaction mode of the document uploading part of the present invention.

[0056] Figure 3 The following is a flowchart of an exemplary document interaction interface generation of the present invention.

[0057] Figure 4 The figure is a flowchart of an exemplary document interaction dialogue of the present invention.

[0058] Figure 5 The figure is a schematic diagram of the module composition of an exemplary document intelligent interaction system of the present invention. DETAILED DESCRIPTION

[0059] The embodiments disclosed in the present invention are described in more detail below with reference to the accompanying drawings. Although certain embodiments disclosed in the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as being limited to the embodiments described herein, which are instead provided for a more thorough and complete understanding of the present invention. It should be understood that the drawings and embodiments disclosed in the present invention are only for exemplary purposes and are not intended to limit the scope of protection disclosed in the present invention.

[0060] In the description of the embodiments disclosed in the present invention, if there are terms, the term "including" and similar terms should be understood as open inclusion, that is, "including but not limited to". The term "based on" should be understood as "based at least in part on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The terms "first", "second", etc. can refer to different or the same objects. Other explicit and implicit definitions may also be included below.

[0061] Taking PDF reading tools as an example, there are currently a variety of solutions for online preview of PDF documents on the market, mainly including: 1) General PDF readers: such as Adobe Acrobat Reader, Foxit Reader, etc. These software provide basic PDF browsing and editing functions, but are mainly desktop applications. 2) Online PDF preview tools: such as cloud services such as Google Docs and Microsoft OneDrive, which allow users to preview and edit PDF documents online, but the functions are relatively limited. 3) Browser plug-ins: such as open source projects such as PDF.js and kkfileview, which preview PDF documents directly through the browser and are suitable for web page embedding, but lack personalized customization capabilities. In general, the interface of traditional PDF reading tools is outdated, and they only provide basic reading functions, which is not convenient for expanding personalized function options.

[0062] In order to solve the above problems existing in the prior art PDF document reading, see Figure 1 In a first aspect, the present invention provides a method for intelligent interaction of documents. Of course, in the present invention, the type of the document includes but is not limited to a PDF document. The method for intelligent interaction of the document includes:

[0063] S100, receiving documents uploaded by users and generating a document list; in the present invention, the document uploading methods include local document uploading and URL link uploading; the page will combine user rights and interests, give priority to verifying the legitimacy of the document, and after passing the verification, the large file to be uploaded will be divided into multiple smaller fragments, making full use of network bandwidth and server resources, and uploading them to OSS for unified storage management.

[0064] Specifically, if the document uploading method is local document uploading, the receiving of the document uploaded by the user and the generating of the document list includes: receiving the document selected by the user based on clicking the upload local document option;

[0065] The document is checked for legality and compliance and user rights are verified; when the legality and compliance check and the user rights verification are passed, the document is sliced ​​and uploaded to the OSS system for unified storage management, and a document list is generated.

[0066] If the document is uploaded by uploading a URL link, the process of receiving the document uploaded by the user and generating a document list includes: receiving the URL link uploaded by the user; performing a legal compliance check on the URL link and performing a user rights verification; after the legal compliance check and the user rights verification pass, downloading the document based on the URL link and performing a document format and document size verification; after the document format and the document size verification pass, slicing the document and uploading it to the OSS system for unified storage management, and generating a document list. The complete document upload process of the intelligent interaction method of the document of the present invention can be as follows: Figure 2 As shown,

[0067] S200, receiving a user's reading request for a target document in the document list, generating a conversational reading interactive interface, the conversational reading interactive interface including a tool area, a document outline area, a document current reading area, and a conversation window area; in the present invention, a user creates a session through a URL or locally uploaded document, and performs session management. When a user selects or switches a document, the document link of the target document is automatically obtained, and the PDF content is read using the pdf-dist plug-in. The document is displayed in a customized UI interface, for example, supporting functions such as generating and viewing thumbnails and outlines, content search, page switching, and zooming.

[0068] In an optional implementation of step S200 of the first aspect of the present invention, the receiving a user's request to read the target document in the document list and generating a conversational reading interaction interface includes: receiving a user's request to read the target document in the document list, obtaining a document link of the target document and initialization data of the conversational reading interaction interface; generating a partition framework of the conversational reading interaction interface based on the initialization data; generating a tool set in a tool area of ​​the partition framework; obtaining a directory of the target document based on the link; generating a hierarchically foldable and expandable outline tree in a document outline area of ​​the partition framework based on the directory; and displaying the homepage of the target document in the current reading area of ​​the document.

[0069] Specifically, the process of generating a conversational reading interaction interface is as follows: First, monitor the document clicks, obtain the document link of the target session based on the document clicks, and initialize the PDF viewer and segment loading parameters. Generate a customized static PDF preview interface, small window view, and toolbar. Render the small amount of document content parsed in the visible area to the page node, generate and mount text nodes, and dynamically load more document pages when scrolling the page. After the page initialization is completed, obtain the page number and generate a placeholder thumbnail. When scrolling, dynamically convert the document data into a thumbnail to replace the placeholder. Generate outline data and display it in combination with a tree control to support hierarchical expansion or folding. When the user switches pages or searches for content, dynamically render the corresponding document content, scroll to the visible area, and highlight the matching text when searching. When the user adjusts the zoom ratio, perform corresponding zoom operations on the main content and document nodes. When the page is closed or destroyed, cancel the corresponding monitoring task. The complete steps for generating a conversational reading interaction interface according to the present invention can be as follows Figure 3 shown.

[0070] S300. Generate and display a document summary of the target document in the conversation window area through a document assistant, wherein the document assistant includes a ChatGPT natural language processing model. In the present invention, while rendering the document content, a corresponding document assistant is created using ChatGPT to generate a document summary that is displayed in the chat area on the right side of the page. Users can use the corresponding prompts to dynamically summarize chapters / knowledge points / cases, extract keywords, extend knowledge, etc. through chatting.

[0071] In an optional implementation of step S300 of the first aspect of the present invention, generating and displaying the document summary of the target document in the session window area by means of a document assistant includes: determining whether the page displayed in the current reading area of ​​the document is the homepage of the target document; if the current page displayed in the current reading area of ​​the document is the homepage of the target document, obtaining the summary page of the target document by means of a document assistant, and displaying the full-text summary of the target document in the session window area; if the current page displayed in the current reading area of ​​the document is not the homepage of the target document, reading the content of the current page displayed in the current reading area of ​​the document and understanding its semantics by means of a document assistant, obtaining a page summary of the current page, and displaying the page summary in the session window area.

[0072] In the present invention, after the tool area, document outline area, and document current reading area are loaded, the system will create a corresponding ChatGPT document assistant, create a thread to send the corresponding summary prompt, and generate a document summary to display in the chat area (i.e., the session window area). If the user is not satisfied with the content of the summary page summary or page summary generated for the first time, the user can also click the regenerate summary button to generate it again, or chat with AI to obtain more document-related knowledge points.

[0073] S400, real-time monitoring of the user's content selection of the target document in the current reading area of ​​the document and obtaining the user's question; in an optional implementation of step S400 of the first aspect of the present invention, the real-time monitoring of the user's content selection of the target document in the current reading area of ​​the document and obtaining the user's question includes: obtaining the cursor style of the current reading area of ​​the document; judging whether the cursor style is a text selection cursor; when the cursor style is the text selection cursor, binding a highlight selection function to the text selection cursor; real-time monitoring of the moving path of the text selection cursor; when the text selection cursor selects text, converting the text symbols passed by the moving path into a highlight display style; obtaining a text set with a state of the highlight display style and generating a question input pop-up window; receiving the user's input question through the question input pop-up window. In the present invention, a quick-connect instruction function can also be generated at the top of the question input pop-up window, and the quick-connect instruction function includes but is not limited to listing keywords, extracting main ideas, extracting cases, summarizing chapters, generating research reviews, and proposing future development directions. After the user clicks the corresponding quick-connect instruction function, the document assistant will send the corresponding answer generated based on the full text.

[0074] S500, generating an answer to the question through the document assistant and displaying it in the conversation window area. In an optional implementation of the first aspect of the present invention, generating an answer to the question through the document assistant and displaying it in the conversation window area includes: extracting keywords from the text set of the question through the document assistant; extending the knowledge of the text set based on the keywords; obtaining the semantics of the question, and obtaining the answer to the question from the knowledge extension based on the semantics; and binding the text set of the question and the answer to the question and displaying them in the conversation window area.

[0075] In an optional implementation of the first aspect of the present invention, after the text set of the question and the answer are bound and displayed in the conversation window area, it also includes: obtaining a number of related questions from the historical question records of the text set or the question association library through the document assistant; displaying the several related questions below the answer for the user to click on and obtain a more in-depth answer through the document assistant or. In the present invention, exemplarily, below each answer, we will list three related questions to provide more directions for the user's conversation. For more explanations, please click the "Tell me more" button to receive more explanations. When the user wants to change the question that the user has sent and send it again, just click the edit button next to the question, and the question will be copied to the input box, or select the text content of the answer, and select the explanation or quote. The interactive process of the conversation part of the present invention can be as follows Figure 4 shown.

[0076] In other optional embodiments of the first aspect of the present invention, the intelligent interaction method of the document of the present invention can also provide the function of summarizing the document to generate a mind map. Specifically, the user can quickly understand and assist in reading the content of the document through the AI ​​dialogue with the ChatGPT document assistant, and additionally provide the function of generating a mind map. The user clicks to generate a mind map, parses the file content with the help of AI, summarizes the viewpoints of the file according to the prompt words, and outputs it in markdown syntax, and then uses markmap to convert the text into a mind map and render it to the corresponding node of the page for display. The user can zoom, download markdown text or png pictures, regenerate, and other operations to facilitate users to summarize and spread.

[0077] In general, the present invention is based on online reading and intelligent dialogue solutions for documents, helping users to easily read and edit PDF files or links on multiple devices. With the help of AI tools, users can quickly understand and assist in reading PDF content, including chapter, knowledge point and case summary, keyword extraction, and knowledge expansion. Our service aims to improve document processing efficiency and reading experience, and provide users with a more efficient, secure and personalized PDF reading experience.

[0078] See also Figure 5 According to a second aspect of the present invention, there is provided a document intelligent interaction system, the intelligent interaction system comprising:

[0079] The document receiving module 10 is used to receive documents uploaded by users and generate a document list;

[0080] A reading request receiving module 20, for receiving a user's reading request for a target document in the document list, and generating a conversational reading interactive interface, wherein the conversational reading interactive interface includes a tool area, a document outline area, a document current reading area, and a conversation window area;

[0081] A summary generation module 30, configured to generate and display a document summary of the target document in the conversation window area through a document assistant, wherein the document assistant includes a ChatGPT natural language processing model;

[0082] A question acquisition module 40, used to monitor the user's selection of the content of the target document in the current reading area of ​​the document in real time and acquire the user's question;

[0083] The answer display module 50 is used to generate an answer to the question through the document assistant and display it in the conversation window area.

[0084] In an optional implementation of the second aspect of the present invention, the reading request receiving module includes:

[0085] A request receiving unit, configured to receive a user's request to read a target document in the document list, and obtain a document link of the target document and initialization data of the conversational reading interaction interface;

[0086] A framework generating unit, configured to generate a partition framework of the conversational reading interaction interface based on the initialization data;

[0087] A tool set generation unit, used for generating a tool set in a tool area of ​​the partition framework;

[0088] A directory acquisition unit, configured to acquire the directory of the target document based on the link;

[0089] An outline creation unit, for generating a hierarchically foldable and expandable outline tree in a document outline area of ​​the partition framework based on the directory;

[0090] The default display unit is used to display the home page of the target document in the current reading area of ​​the document.

[0091] In an optional implementation of the second aspect of the present invention, the summary generation module includes:

[0092] A homepage determination unit, used to determine whether the page displayed in the current reading area of ​​the document is the homepage of the target document;

[0093] A summary page display unit, configured to obtain a summary page of the target document through a document assistant if the current page displayed in the current reading area of ​​the document is the homepage of the target document, and display the full text summary of the target document in the session window area;

[0094] A page summary display unit is used to read the content and semantically understand the current page displayed in the current reading area of ​​the document through a document assistant if the current page displayed in the current reading area of ​​the document is not the homepage of the target document, obtain a page summary of the current page, and display the page summary in the session window area.

[0095] In an optional implementation of the second aspect of the present invention, the question acquisition module includes:

[0096] A cursor style acquisition unit, used to acquire a cursor style in a current reading area of ​​the document;

[0097] A cursor judging unit, used to judge whether the cursor style is a text selection cursor;

[0098] A cursor function binding unit, used for binding a highlight selection function to the text selection cursor when the cursor style is the text selection cursor;

[0099] A cursor path monitoring unit, used for monitoring the moving path of the text selection cursor in real time;

[0100] A highlight processing unit, used for converting the text symbols passed by the moving path into a highlight display style when the text selection cursor selects a text;

[0101] A text set acquisition unit, used to acquire a text set in the highlighted display style and generate a question input pop-up window;

[0102] The question acquisition unit is used to receive the question input by the user through the question input pop-up window.

[0103] In an optional implementation of the second aspect of the present invention, the answer display module includes:

[0104] A keyword extraction unit, configured to extract keywords from the text set of the question through the document assistant;

[0105] A knowledge extension unit, used for extending the knowledge of the text set based on the keywords;

[0106] An answer acquisition unit, used to acquire the semantics of the question, and acquire the answer to the question from the knowledge extension based on the semantics;

[0107] A conversation display unit is used to bind the text set of the question and the answer and display them in the conversation window area.

[0108] In an optional implementation of the second aspect of the present invention, the answer display module further includes:

[0109] A related question acquisition unit, configured to acquire a number of related questions from the historical question records of the text set or the question association relationship library through the document assistant;

[0110] The related question display unit is used to display a number of the related questions below the answer, so that the user can click on them to obtain a more in-depth answer through the document assistant or.

[0111] In an optional implementation of the second aspect of the present invention, the document uploading method includes local document uploading and URL link uploading;

[0112] If the document uploading method is local document uploading, the receiving of the document uploaded by the user and the generating of the document list include:

[0113] Receive the document selected by the user based on clicking the upload local document option;

[0114] Conduct legal and compliance checks on the documents and verify user rights and interests;

[0115] After the legal compliance detection and the user rights verification are passed, the document is sliced ​​and uploaded to the OSS system for unified storage management, and a document list is generated;

[0116] If the document is uploaded in a URL link upload mode, the step of receiving the document uploaded by the user and generating a document list includes:

[0117] Receive URL links uploaded by users;

[0118] Conduct legal and compliance checks on the URL link and verify user rights;

[0119] After the legal compliance detection and the user rights verification are passed, the document is downloaded based on the URL link, and the document format and document size are verified;

[0120] After the document format and the document size are verified, the document will be sliced ​​and uploaded to the OSS system for unified storage management, and a document list will be generated.

[0121] In addition, although each operation is described in a specific order, this should be understood as requiring such operation to be performed in the specific order shown or in a sequential order, or requiring that all illustrated operations should be performed to obtain desired results. Similarly, although some specific implementation details are included in the above discussion, these should not be interpreted as limiting the scope of the present disclosure. Certain features described in the context of a separate embodiment can also be implemented in a single implementation in combination. On the contrary, the various features described in the context of a single implementation can also be implemented in multiple implementations individually or in any suitable sub-combination.

[0122] Although the subject matter has been described in language specific to structural features and / or methodological logical actions, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. On the contrary, the specific features and actions described above are merely example forms of implementing the claims.

Claims

1. A method for intelligent interaction of documents, characterized in that: include: Receive documents uploaded by users and generate a document list; Receiving a user's reading request for a target document in the document list, and generating a conversational reading interaction interface, wherein the conversational reading interaction interface includes a tool area, a document outline area, a document current reading area, and a conversation window area; Generate and display a document summary of the target document in the conversation window area by means of a document assistant, wherein the document assistant includes a ChatGPT natural language processing model; Real-time monitoring of the user's selection of the content of the target document in the current reading area of ​​the document and obtaining the user's questions; The answer to the question is generated by the document assistant and displayed in the conversation window area.

2. The intelligent document interaction method according to claim 1, characterized in that: The receiving a user's request to read a target document in the document list and generating a conversational reading interaction interface comprises: Receiving a user's request to read a target document in the document list, and obtaining a document link of the target document and initialization data of the conversational reading interaction interface; Generate a partition framework of the conversational reading interaction interface based on the initialization data; generating a tool set in a tool area of ​​the partition framework; Obtaining a thumbnail directory of the target document based on the link; Generate a hierarchically foldable and expandable outline tree in the document outline area of ​​the partition framework based on the directory; The first page of the target document is displayed in the current reading area of ​​the document.

3. The intelligent document interaction method according to claim 1, characterized in that: The step of generating and displaying a document summary of the target document in the conversation window area by means of a document assistant includes: Determine whether the page displayed in the current reading area of ​​the document is the homepage of the target document; If the current page displayed in the current reading area of ​​the document is the homepage of the target document, the summary page of the target document is obtained through the document assistant, and the full text summary of the target document is displayed in the session window area; If the current page displayed in the current reading area of ​​the document is not the homepage of the target document, the document assistant reads the content and semantically understands the current page displayed in the current reading area of ​​the document, obtains a page summary of the current page, and displays the page summary in the session window area.

4. The intelligent document interaction method according to claim 1, characterized in that: The real-time monitoring of the user's selection of the content of the target document in the current reading area of ​​the document and obtaining the user's question comprises: Get the cursor style of the current reading area of ​​the document; Determine whether the cursor style is a text selection cursor; When the cursor style is the text selection cursor, a highlight selection function is bound to the text selection cursor; Performing real-time monitoring on the moving path of the text selection cursor; When the text selection cursor is selecting a word, the word symbol that the moving path passes through is converted into a highlight display style; Obtain a text set in the highlighted display style and generate a question input pop-up window; The question input pop-up window is used to receive questions input by the user.

5. The intelligent document interaction method according to claim 4, characterized in that: The generating of the answer to the question by the document assistant and displaying it in the conversation window area includes: Extracting keywords from the text set of the question by the document assistant; Performing knowledge extension on the text set based on the keywords; Acquire the semantics of the question, and obtain an answer to the question from the knowledge extension based on the semantics; The text set of the question and the answer are bound and displayed in the conversation window area.

6. The intelligent document interaction method according to claim 5, characterized in that: After the text set of the question and the answer are bound and displayed in the conversation window area, the following further comprises: Acquiring, by the document assistant, a number of related questions from the historical question records of the text set or the question association relationship library; A number of the related questions are displayed below the answer, so that the user can click on them to obtain a more in-depth answer through the document assistant or other means.

7. The intelligent document interaction method according to claim 1, characterized in that: The document uploading method includes local document uploading and URL link uploading; If the document uploading method is local document uploading, the receiving of the document uploaded by the user and the generating of the document list include: Receive the document selected by the user based on clicking the upload local document option; Conduct legal and compliance checks on the documents and verify user rights and interests; After the legal compliance detection and the user rights verification are passed, the document is sliced ​​and uploaded to the OSS system for unified storage management, and a document list is generated; If the document is uploaded in a URL link upload mode, the step of receiving the document uploaded by the user and generating a document list includes: Receive URL links uploaded by users; Conduct legal and compliance checks on the URL link and verify user rights; After the legal compliance detection and the user rights verification are passed, the document is downloaded based on the URL link, and the document format and document size are verified; After the document format and the document size are verified, the document will be sliced ​​and uploaded to the OSS system for unified storage management, and a document list will be generated.

8. An intelligent document interaction system, characterized in that: The intelligent interaction system comprises: The document receiving module is used to receive documents uploaded by users and generate a document list; A reading request receiving module, used to receive a user's reading request for a target document in the document list, and generate a conversational reading interaction interface, wherein the conversational reading interaction interface includes a tool area, a document outline area, a document current reading area, and a conversation window area; A summary generation module, used to generate and display a document summary of the target document in the conversation window area through a document assistant, wherein the document assistant includes a ChatGPT natural language processing model; A question acquisition module, used to monitor the user's selection of the content of the target document in the current reading area of ​​the document in real time and acquire the user's question; The answer display module is used to generate the answer to the question through the document assistant and display it in the conversation window area.

Images