Intelligent medical literature searching and question-answering system
By designing an intelligent medical literature search question and answer system, using natural language processing and deep learning technology, the problem that existing systems are difficult to understand user intentions and provide accurate search results is solved, efficient literature search and automatic information summary are achieved, and user experience is improved.
Patent Information
- Application Number
- CN202411580743.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-07
- Publication Date
- 2025-02-25
AI Technical Summary
The existing medical literature search system is difficult to accurately understand the user's query intention, the accuracy and relevance of the search results are limited, and the functions of automatic summary and dialogue outline generation are lacking, resulting in users still spending a lot of time reading and organizing after obtaining the information.
An intelligent medical literature search question and answer system is designed, including a user interface module, a multi-round dialogue processing module, a literature search module, a document content analysis module, an abstract generation module, a dialogue outline generation module and a performance tracking module. Through natural language processing and deep learning technology, the understanding of user query intentions, efficient literature search, automatic summary of content and real-time generation of dialogue outlines is achieved.
It realizes an accurate understanding of user query intentions, provides more accurate literature search results and professional suggestions, and improves the efficiency of literature search and user information acquisition experience.
Smart Images
Figure 00000000_0000_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of artificial intelligence technology, and in particular relates to an intelligent medical literature search question-answering system. Background Art
[0002] With the development of the medical industry and the advancement of information technology, the amount of medical literature has grown rapidly, providing researchers and medical professionals with a wealth of knowledge resources. However, faced with a huge amount of literature information, how to quickly and accurately find the required information and understand its content has become a challenge. Traditional literature retrieval methods are often time-consuming and inefficient, and it is difficult to meet users' needs for personalized and real-time information processing. In order to improve the efficiency and accuracy of medical literature retrieval, an intelligent literature search question-answering system is urgently needed.
[0003] Existing medical literature retrieval systems mostly rely on keyword searches and lack an in-depth understanding of user query intent, resulting in limited accuracy and relevance of retrieval results. In addition, these systems usually do not have the function of automatically summarizing document content and generating dialogue outlines, so users still need to spend a lot of time reading and organizing information after obtaining it. Therefore, developing a medical literature search question-answering system that can understand complex query intent, automatically retrieve and summarize document content, and update dialogue outlines in real time is of great significance to improving the efficiency of medical research and clinical practice. Summary of the invention
[0004] In order to solve the above technical problems, the present invention proposes an intelligent medical literature search question-answering system to solve the above problems existing in the prior art.
[0005] To achieve the above object, the present invention provides an intelligent medical literature search question-answering system, comprising:
[0006] A user interface module, used to obtain a user's query request;
[0007] A multi-round dialogue processing module, used to understand the user's query request through multi-round interactions and generate the user's query intention;
[0008] A document retrieval module is used to retrieve relevant academic documents in a medical document database based on the user's query intention;
[0009] A document content analysis module, used to extract key information of the academic document;
[0010] A summary generation module, used to generate a summary of the academic document based on the key information, and feed back the summary of the academic document to the user;
[0011] A conversation outline generation module, used to generate a conversation outline according to the user's query history, and to help the user track and organize the acquired information based on the conversation outline;
[0012] The performance tracking module is used to monitor the execution performance of each functional module of the system.
[0013] Preferably, the user interface module is also used to: generate a command line interface or a graphical user interface, and provide multi-language support.
[0014] Preferably, the multi-round dialogue processing module is also used to: understand user input through natural language processing technology and maintain the conversation context.
[0015] Preferably, the document retrieval module is also used to: perform document retrieval using keyword matching and Boolean logic combination, and simultaneously interface with an external medical document database API to obtain the latest academic documents.
[0016] Preferably, the document content parsing module includes:
[0017] HTML content parser, used to extract text content from HTML format of documents;
[0018] Text mining tools to identify and extract key medical information from text.
[0019] Preferably, the summary generation module is also used to: analyze the key information through deep learning technology to generate a summary of the academic document.
[0020] Preferably, the dialogue outline generating module further includes:
[0021] Conversation history analyzer, used to analyze the user's query history;
[0022] An outline builder to build a conversation outline based on the analysis results.
[0023] Preferably, the performance tracking module includes:
[0024] Timer, used to measure the response time of each module in the system;
[0025] Performance logger, used to record performance data of system operations.
[0026] Compared with the prior art, the present invention has the following advantages and technical effects:
[0027] The present invention provides a medical literature search question-answering system, which aims to achieve accurate understanding of user query intent, efficient retrieval of medical literature, automatic summary of literature content, and real-time generation of dialogue outlines through artificial intelligence technology, especially natural language processing and deep learning technology. The system interacts with users through a multi-round dialogue mechanism, gradually refines and confirms user needs, and provides more accurate literature search results and professional suggestions. BRIEF DESCRIPTION OF THE DRAWINGS
[0028] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:
[0029] Figure 1 Schematic diagram of an intelligent medical literature search question-answering system according to an embodiment of the present invention. DETAILED DESCRIPTION
[0030] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.
[0031] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.
[0032] Embodiment 1
[0033] like Figure 1 As shown, this embodiment provides an intelligent medical literature search question-answering system, including:
[0034] A user interface module, used to obtain a user's query request;
[0035] A multi-round dialogue processing module, used to understand the user's query request through multi-round interactions and generate the user's query intention;
[0036] A document retrieval module is used to retrieve relevant academic documents in a medical document database based on the user's query intention;
[0037] A document content analysis module, used to extract key information of the academic document;
[0038] A summary generation module, used to generate a summary of the academic document based on the key information, and feed back the summary of the academic document to the user;
[0039] A conversation outline generation module, used to generate a conversation outline according to the user's query history, and to help the user track and organize the acquired information based on the conversation outline;
[0040] The performance tracking module is used to monitor the execution performance of each functional module of the system.
[0041] According to a further optimization scheme, the user interface module is also used to generate a command line interface or a graphical user interface and provide multi-language support.
[0042] To further optimize the solution, the multi-round dialogue processing module is also used to: understand user input through natural language processing technology and maintain the conversation context.
[0043] To further optimize the solution, the document retrieval module is also used to: perform document retrieval using keyword matching and Boolean logic combination, and simultaneously interface with an external medical literature database API to obtain the latest academic literature.
[0044] Further optimizing the scheme, the document content parsing module includes:
[0045] HTML content parser, used to extract text content from HTML format of documents;
[0046] Text mining tools to identify and extract key medical information from text.
[0047] To further optimize the solution, the summary generation module is also used to: analyze the key information through deep learning technology to generate a summary of the academic document.
[0048] To further optimize the solution, the dialogue outline generation module further includes:
[0049] Conversation history analyzer, used to analyze the user's query history;
[0050] An outline builder to build a conversation outline based on the analysis results.
[0051] Further optimizing the solution, the performance tracking module includes:
[0052] Timer, used to measure the response time of each module in the system;
[0053] Performance logger, used to record performance data of system operations.
[0054] This embodiment also includes an operation method of the medical literature search question-answering system, including:
[0055] a. Receive the query request input by the user;
[0056] b. Use the multi-round dialogue processing module to understand the user's query intention;
[0057] c. Based on the understood query intent, retrieve relevant academic literature from the medical literature database through the literature retrieval module;
[0058] d. Use the document content analysis module to analyze the retrieved academic documents and extract key information;
[0059] e. Generate abstracts of academic documents through the abstract generation module;
[0060] f. Generate a dialogue outline based on the user's query history using a dialogue outline generation module;
[0061] g. Monitor the execution performance of each functional module of the system through the performance tracking module.
[0062] The step of receiving a query request inputted by a user includes receiving inputted through a command line interface or a graphical user interface.
[0063] The step of understanding the user's query intention using the multi-turn dialogue processing module includes using natural language processing technology to understand user input and maintain the conversation context.
[0064] The steps for retrieving relevant academic literature from the medical literature database through the literature retrieval module include:
[0065] a. Use keyword matching and Boolean logic combination to conduct literature search;
[0066] b. Obtain the latest academic literature through API interface with external medical literature database.
[0067] The steps of parsing the retrieved academic documents using the document content parsing module include:
[0068] a. Use HTML content parser to extract text content from HTML format of document;
[0069] b. Use text mining tools to identify and extract key medical information from text.
[0070] The abstract generation module generates abstracts of academic documents using deep learning technology based on the key content of the document.
[0071] The steps of using the dialogue outline generation module to generate a dialogue outline according to the user's query history include:
[0072] a. Use the conversation history analyzer to analyze the user's query history;
[0073] b. Use the outline builder to build a conversation outline based on the analysis results.
[0074] The steps to monitor the execution performance of each functional module of the system through the performance tracking module include:
[0075] a. Use a timer to measure the response time of each module in the system;
[0076] b. Use the performance logger to record performance data of system operations.
[0077] This embodiment also provides a computer-readable storage medium, which is a computer program for implementing the operating method of the medical literature search question-and-answer system.
[0078] This embodiment also provides an electronic device, including a processor and a memory, wherein the memory stores a computer program for implementing the operating method of the medical literature search question and answer system, and the processor executes the program to implement the operating method of the medical literature search question and answer system.
[0079] This embodiment also provides a data synchronization method for a medical literature search question-and-answer system, including a data synchronization step between an electronic device and a server to ensure that a user's health records and diagnostic data remain consistent across multiple devices.
[0080] This embodiment also provides a data backup and recovery method for a medical literature search question-and-answer system, which includes a server regularly backing up user data and providing a data recovery function when necessary.
[0081] This embodiment also provides a data access log and audit method for a medical literature search question-and-answer system. The method includes a server recording an access log of user data and providing an audit function to ensure compliance of data access.
[0082] This embodiment also provides an API key configuration method for a medical literature search question-and-answer system, which includes a user obtaining an API key from OpenAI and MediSearch and configuring the key in the software to enable access to OpenAI services.
[0083] This embodiment also provides a method for installing a medical literature search question-answering system, the method comprising the following steps:
[0084] a. Install the Python interpreter;
[0085] b. Install the required Python libraries through the package management tool;
[0086] c. Configure API key.
[0087] This embodiment also provides a detailed description of the use of the medical literature search question-answering system, which includes the following contents:
[0088] a. Provide advice on medical related issues;
[0089] b. Generate a dialogue outline;
[0090] c. Handle multiple rounds of dialogue;
[0091] d. Reference and link processing.
[0092] This embodiment also provides a system architecture of a medical literature search question-answering system, which includes:
[0093] a. Client server model;
[0094] b. Command line interface as front end;
[0095] c. The backend includes data processing module, AI model calling module and storage module.
[0096] This embodiment also provides a key technology and tool of a medical literature search question-answering system, including:
[0097] a. Python programming language;
[0098] b. Dependent libraries such as `httpx`, `beautifulsoup4`, `uuid`, and `openai`.
[0099] This embodiment also provides the performance and scalability of a medical literature search question-and-answer system. The system design takes into account high-concurrency processing and low-latency response to support real-time user interaction and horizontal expansion.
[0100] This embodiment also provides a method for generating academic paper summaries and dialogue outlines in a medical literature search question-answering system, the method comprising:
[0101] a. Use the GPT model to parse academic papers and HTML content;
[0102] b. Summarize key information and produce high-quality summaries;
[0103] c. Update the conversation outline in real time based on the user’s questions and conversation history.
[0104] This embodiment also provides a medical information retrieval and suggestion generation method of a medical literature search question-answering system, the method comprising:
[0105] a. Use the MediSearch database to search for relevant medical literature;
[0106] b. Combine OpenAI models to generate accurate medical advice.
[0107] Embodiment 2
[0108] This embodiment provides an intelligent medical literature search question-answering system, including:
[0109] 1. User interface module:
[0110] Design an intuitive and user-friendly graphical user interface (GUI), including a search bar, history, results display area, and user preferences.
[0111] Implement a command line interface (CLI) to provide shortcuts and scriptable access for advanced users.
[0112] The interface supports touch operation and voice input, and adapts to mobile devices and barrier-free access requirements.
[0113] Provide a multi-language interface, users can switch languages as needed, including but not limited to Chinese, English, Spanish, etc.
[0114] 2. Multi-round dialogue processing module:
[0115] Leverage natural language processing (NLP) technologies, including semantic analysis, context understanding, and intent recognition, to maintain the coherence of multi-round conversations.
[0116] Implement a dialog manager to maintain conversation state and history, ensuring that the system can reference previous interactions to provide relevant responses.
[0117] Using machine learning algorithms, we continuously optimize conversation strategies and personalized recommendations based on user feedback and behavior patterns.
[0118] 3. Literature search module:
[0119] Build a comprehensive medical literature database index that supports multiple search methods, including subject terms, authors, publication dates, and journal names.
[0120] Implement API integration with external databases such as PubMed, EMBASE, Cochrane Library, etc. to access the latest research results.
[0121] Design an advanced search interface that allows users to refine search criteria using Boolean logic (AND, OR, NOT) and field-specific searches.
[0122] 4. Document content analysis module:
[0123] Develop an HTML content parser, using BeautifulSoup or another parsing library to extract and clean the text content from the document.
[0124] Implement text mining tools to extract key information from documents using keyword extraction, named entity recognition (NER), and concept relationship mapping.
[0125] 5. Summary generation module:
[0126] Use Transformer-based deep learning models, such as BERT or GPT, to generate summaries of document content.
[0127] Provides users with the option to customize the length and level of detail of the summary, as well as the style of the generated summary (e.g., scholarly or popular).
[0128] Implement a summary quality assessment mechanism to ensure that the generated summary accurately reflects the core content and conclusions of the original text.
[0129] 6. Dialogue outline generation module:
[0130] Develop a conversation analyzer that uses text analysis techniques to identify key topics and concepts in conversations.
[0131] Implement an outline builder to automatically organize the conversation content into a structured outline, allowing users to quickly grasp the key points of the conversation.
[0132] Provide a visual representation of the outline, such as a mind map or timeline, to enhance readability and retention of the information.
[0133] 7. Performance tracking module:
[0134] Implement a performance monitoring system to track and record the response time and processing capacity of each module of the system.
[0135] Design a log management system to collect and analyze system operation logs, including performance indicators and abnormal events.
[0136] Provides a performance optimization tool that automatically adjusts system configuration and resource allocation based on performance data.
[0137] 8. System architecture design:
[0138] Adopt microservice architecture to decompose the system into independent service units to improve the maintainability and scalability of the system.
[0139] Implement containerized deployment and use tools such as Docker and Kubernetes to manage and scale services.
[0140] Design a high availability architecture and use load balancing and redundancy to ensure stable operation of the system.
[0141] 9. Data synchronization and backup:
[0142] Design a data synchronization protocol to ensure user data consistency across different devices and sessions.
[0143] Implement a data backup framework that supports regular full backup and real-time incremental backup, as well as off-site backup strategies.
[0144] 10. Audit and Compliance:
[0145] Implement an audit trail system that records all user actions and system events to meet compliance requirements.
[0146] Design a compliance checking tool to periodically verify that system operations meet healthcare industry regulations and standards.
[0147] Embodiment 3
[0148] This embodiment provides an intelligent medical literature search question-answering system, including:
[0149] Hardware requirements
[0150] You will need a modern computer with a stable internet connection and the ability to connect to the OpenAI service.
[0151] Software requirements
[0152] Operating system: Windows, macOS or Linux.
[0153] Programming environment: Python 3.7 or higher, ensuring compatibility with the latest libraries and APIs.
[0154] Dependent libraries: The software depends on several Python libraries, including httpx for network requests, beautifulsoup4 (ie bs4) for HTML content parsing, uuid for generating unique session IDs, and the openai library to call OpenAI's AI model.
[0155] API key: Users need to obtain API keys from OpenAI and MediSearch and set them correctly in the software configuration to access OpenAI services.
[0156] Installation Guide
[0157] 1. Install Python:
[0158] Visit the Python official website ([python.org](https: / / python.org)), download and install Python 3.7 or higher. During the installation, make sure to select the "Add Python to PATH" option.
[0159] 2. Install dependent libraries:
[0160] Open a command prompt or terminal and execute the following command to install the required Python libraries:
[0161] ```
[0162] pipinstallhttpxbeautifulsoup4openaiuuid
[0163] ```
[0164] These libraries will support functions such as network requests, HTML content parsing, and AI model calls.
[0165] 3. Configure API key:
[0166] After obtaining API keys from OpenAI and MediSearch, you need to configure them in the software.
[0167] Quick Start
[0168] 1. Start the software:
[0169] Starting the software through the command line interface usually requires running a Python script, for example:
[0170] ```
[0171] pythonSearch.py
[0172] ```
[0173] Follow the on-screen prompts.
[0174] 2. Enter your query:
[0175] When prompted, type your query question or keywords and press Enter to submit.
[0176] 3. View the results:
[0177] The software will process your query and display relevant suggestions and information summaries.
[0178] 4. End the session:
[0179] Following the prompt, you can end the session by entering a specific command such as "exit".
[0180] Also includes:
[0181] 1. Provide advice based on medical related issues:
[0182] Enter the query content, and the software will automatically obtain the answer to the corresponding question through the API, and provide corresponding references and their links.
[0183] 2. Generate a dialogue outline:
[0184] During the conversation, the software will use the `generate_outline` function to update the conversation outline in real time based on the user's query history, helping the user track the progress and key points of the entire conversation.
[0185] 3. Handling multi-round conversations:
[0186] Users can enter multiple queries in succession, and the software will process and provide feedback one by one. After each query, the conversation outline will be updated to reflect new information and suggestions.
[0187] 4. References and links processing:
[0188] If the query results contain external links or references, the software will provide a summary and, if possible, parse the main content of these resources.
[0189] This embodiment is built on a client-server model, where users interact with the system through a command line interface (CLI), and the system backend is responsible for processing logic and data storage.
[0190] Front end: command line interface, responsible for receiving user input and displaying processing results.
[0191] Backend: includes data processing module, AI model calling module and storage module.
[0192] Data processing module: responsible for data reception, analysis and preprocessing.
[0193] AI model calling module: interacts with OpenAI API and MediSearch API to perform content summarization, dialogue outline generation and professional medical information retrieval.
[0194] Storage module: stores user session data and historical interaction records for context management and historical data analysis.
[0195] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
Claims
1. An intelligent medical literature search question-answering system, characterized in that: include: A user interface module, used to obtain a user's query request; A multi-round dialogue processing module, used to understand the user's query request through multi-round interactions and generate the user's query intention; A document retrieval module is used to retrieve relevant academic documents in a medical document database based on the user's query intention; A document content analysis module, used to extract key information of the academic document; A summary generation module, used for generating a summary of the academic document based on the key information, and feeding back the summary of the academic document to the user; A conversation outline generation module, used to generate a conversation outline according to the user's query history, and to help the user track and organize the acquired information based on the conversation outline; The performance tracking module is used to monitor the execution performance of each functional module of the system.
2. The intelligent medical literature search question-answering system according to claim 1, characterized in that: The user interface module is also used to generate a command line interface or a graphical user interface and provide multi-language support.
3. The intelligent medical literature search question-answering system according to claim 1, characterized in that: The multi-round dialogue processing module is also used to understand user input through natural language processing technology and maintain the conversation context.
4. The intelligent medical literature search question-answering system according to claim 1, characterized in that: The document search module is also used to: perform document search using keyword matching and Boolean logic combination, and simultaneously interface with an external medical document database API to obtain the latest academic documents.
5. The intelligent medical literature search question-answering system according to claim 1, characterized in that: The document content parsing module includes: HTML content parser, used to extract text content from HTML format of documents; Text mining tools to identify and extract key medical information from text.
6. The intelligent medical literature search question-answering system according to claim 1, characterized in that: The summary generation module is also used to: analyze the key information through deep learning technology to generate a summary of the academic document.
7. The intelligent medical literature search question-answering system according to claim 1, characterized in that: The dialogue outline generation module also includes: Conversation history analyzer, used to analyze the user's query history; An outline builder to build a conversation outline based on the analysis results.
8. The intelligent medical literature search question-answering system according to claim 1, characterized in that: The performance tracking module includes: Timer, used to measure the response time of each module in the system; Performance logger, used to record performance data of system operations.