Intelligent knowledge point extracting and labeling method and system for medical textbook

Generate key test points excerpts of medical textbooks through vector representation and semantic similarity evaluation, and combine the medical concept dictionary for intelligent annotation and flash card generation, which solves the problems of low efficiency and poor accuracy of knowledge point annotation in medical textbooks, and realizes personalization Learning experience and efficient exam preparation.

CN118982067AInactive Publication Date: 2024-11-19SUZHOU INST OF SYST MEDICINE

Patent Information

Application Number
CN202411201816.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-29
Publication Date
2024-11-19
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The labeling of knowledge points in medical textbooks relies on manual labor, has low efficiency and poor accuracy, lacks personalized and intelligent auxiliary tools, and it is difficult to fully cover the learning of medical knowledge system.

Method used

By collecting medical examination real questions and textbook resources, using vector representation and semantic similarity assessment, generating key test points excerpts, and intelligently annotate them in combination with the medical concept dictionary, the flash card content is designed to be presented in the form of questions and answers.

Benefits of technology

It improves information accuracy and learning efficiency, reduces manual dependence, provides a personalized learning experience, enhances the convenience of knowledge memory and inspection, and significantly improves the quality and efficiency of exam preparation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118982067A_ABST
    Figure CN118982067A_ABST
Patent Text Reader

Abstract

The invention provides an intelligent knowledge point extracting and labeling method and system for a medical textbook, and relates to the technical field of medical education.The method comprises the steps that medical examination real question resources and textbook resources are collected, medical examination real question texts and textbook texts are converted into vector representation, and the semantic similarity between text fragments is evaluated, so that the medical examination real question resources and the textbook resources are extracted and labeled. Screening out a plurality of medical textbook text fragments which have the highest semantic similarity with the medical examination real question text, and generating a textbook key examination point picking; on the basis of the constructed medical concept dictionary, marking medical concepts in the key examination point books of the textbook, and marking text segments associated with real questions in the key examination point books of the textbook; and designing a flash card generation instruction, generating flash card content in a set format according to the key examination point of the textbook, and presenting the flash card content in a form containing questions and answers through rendering. The learning efficiency of the user is improved, and the accuracy and pertinence of the learning content are ensured.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of medical education technology, and in particular to a method and system for intelligently extracting and annotating knowledge points in medical textbooks. Background Art

[0002] Medical education is highly dependent on the extraction, marking, understanding and memorization of knowledge points. Medical students usually follow three main steps when studying professional courses such as internal medicine and surgery: reading and manually marking important knowledge points, reciting and memorizing these knowledge points, and testing their mastery of knowledge points by doing exercises. However, medical textbooks are huge, and medical students face many challenges when preparing for key medical exams, such as short preparation time, difficulty grasping the key points, low memory efficiency and weak memory of knowledge points.

[0003] At present, the annotation of important knowledge points in medical textbooks mainly relies on manual reading and annotation by medical experts or professional students. For example, the series of books "Notes on Lectures at Peking Union Medical College" and "Lecture Notes on Comprehensive Competence in Western Clinical Medicine" are all based on the excerpts and compilation of expert experience, but these methods have limited use of information technology and artificial intelligence.

[0004] With the development of information technology and artificial intelligence technology, some auxiliary tools have emerged, such as flashcards or Anki memory cards, which improve the efficiency of knowledge memory through electronic and digital means. However, the production of these tools still requires a lot of manual input, and they mainly solve the problem of knowledge memory efficiency, and fail to fully cover the entire system of medical knowledge system learning.

[0005] Large language model technologies, such as ChatGPT, have begun to be explored and applied in the field of education. For example, AnkiBrain products from AnkiWeb and Mapify products from Xmind use the power of large language models to improve the efficiency of making memory cards and converting texts into mind maps. In addition, AI document reading tools such as ChatPDF and ChatDOC provide users with assistance in document learning and problem solving through real-time interactive dialogue capabilities.

[0006] The application of large language model technology in the field of education has made some progress, but existing technical solutions are mainly aimed at general fields and lack special optimization for the medical field. In addition, large language models may generate false or erroneous information, which affects users' trust in AI-generated content.

[0007] In view of the limitations of existing technologies and the special needs of medical education, it is necessary to develop a new technical solution that should combine artificial intelligence technology, especially the capabilities of large language models, to provide medical students with a comprehensive, integrated, and intelligent learning assistance system. The system should be able to intelligently extract and annotate key knowledge points in textbooks, generate tools to assist memory, such as smart flash cards, and be closely integrated with learners' reading and exam preparation processes to improve learning efficiency and quality. At the same time, the system needs to solve the information accuracy problems that may arise from large language models to ensure that the medical information provided to users is reliable and accurate. Summary of the invention

[0008] To this end, an embodiment of the present invention provides a method and system for intelligent extraction and annotation of knowledge points for medical textbooks, which is used to solve the problems of low efficiency, strong manual dependence, insufficient personalization, poor information accuracy, etc. in the prior art in medical learning assistance.

[0009] In order to solve the above problems, an embodiment of the present invention provides a method for intelligent extraction and annotation of knowledge points in medical textbooks, the method comprising: Collect medical exam question resources and medical textbook resources, convert the medical exam question texts and medical textbook texts into vector representations, evaluate the semantic similarity between text fragments, select multiple medical textbook text fragments with the highest semantic similarity to the medical exam question texts, and generate textbook excerpts of key test points; Based on the constructed medical concept dictionary, the medical concepts in the textbook key test point book excerpts are annotated, and the text fragments related to the real questions in the textbook key test point book excerpts are annotated; Design flashcard generation instructions, generate flashcard content in a set format based on the textbook key test points and book excerpts, and present the flashcard content in a table form containing questions and answers through rendering.

[0010] Preferably, medical examination question resources and medical textbook resources are collected, the medical examination question texts and medical textbook texts are converted into vector representations, and by evaluating the semantic similarity between text fragments, multiple medical textbook text fragments with the highest semantic similarity to the medical examination question texts are screened out to generate textbook key points excerpts, specifically including: Collect medical exam questions and medical textbooks; Use the text vectorization model to convert medical exam questions and medical textbook texts into vector representations; Storing vectorized medical textbook texts in a vector database; For each real question, a hybrid search strategy of vectorized search and keyword search is used to evaluate the semantic similarity between text fragments, and multiple medical textbook text fragments with the highest semantic similarity to the real medical examination text are selected as candidate book excerpt fragments; The selected candidate excerpts are reviewed by experts to generate the final version of the textbook key points excerpts.

[0011] Preferably, in the process of vectorizing the text of the medical textbook, the text whose length exceeds the processing upper limit of the text vectorization model is segmented before the text vectorization is performed.

[0012] Preferably, the method for constructing the medical concept dictionary is: A medical concept dictionary is obtained by collating authoritative medical terminology resources and medical textbook vocabularies, wherein the authoritative medical terminology resources include the International Classification of Diseases codes, the Human Phenotype Ontology, and commonly used clinical medical terms released by the National Health Commission; the medical textbook vocabularies include internal medicine and surgery.

[0013] Preferably, based on the constructed medical concept dictionary, the medical concepts in the textbook key test point book excerpts are annotated, and the text fragments related to the real questions in the textbook key test point book excerpts are annotated, specifically including: Based on the constructed medical concept dictionary, scan the medical concepts in the textbook key points book excerpts, annotate the important medical concepts in the textbook key points book excerpts, and specially annotate the medical concepts that appear in the medical examination test texts; When scanning text fragments in the textbook key test point excerpts, if a text fragment is found to have related real questions, the front-end technology is used to insert a real question mark at the end of the text fragment and provide a link to the question page.

[0014] Preferably, a flashcard generation instruction is designed to generate flashcard content in a set format according to the textbook key test point excerpt, and the flashcard content is presented in a table form containing questions and answers through rendering, specifically including: Design flashcard generation instructions, generate flashcard content in markdown format based on the textbook key test points and book excerpts, use front-end technology to render markdown data, and present the flashcard content in a table format containing questions and answers.

[0015] Preferably, the front-end technology includes HTML and JavaScript.

[0016] The embodiment of the present invention further provides a system for intelligent extraction and annotation of knowledge points for medical textbooks. The system is used to implement the above-mentioned method for intelligent extraction and annotation of knowledge points for medical textbooks, and specifically includes: Intelligent book excerpt extraction algorithm module, used to collect medical examination question resources and medical textbook resources, convert medical examination question text and medical textbook text into vector representation, and select multiple medical textbook text fragments with the highest semantic similarity to medical examination question text by evaluating the semantic similarity between text fragments, and generate textbook excerpts of key examination points; Intelligent book excerpt annotation algorithm module, which is used to annotate the medical concepts in the textbook key test point book excerpts based on the constructed medical concept dictionary, and annotate the text fragments related to the real questions in the textbook key test point book excerpts; The intelligent flashcard generation algorithm module is used to design flashcard generation instructions, generate flashcard content in a set format according to the textbook key test points and book excerpts, and present the flashcard content in a table form containing questions and answers through rendering.

[0017] An embodiment of the present invention also provides an electronic device, which includes a processor, a memory and a bus system, wherein the processor and the memory are connected via the bus system, the memory is used to store instructions, and the processor is used to execute the instructions stored in the memory to implement the above-mentioned intelligent extraction and annotation method of knowledge points for medical textbooks.

[0018] An embodiment of the present invention further provides a computer storage medium storing a computer software product. The computer software product includes several instructions for enabling a computer device to execute the above-mentioned intelligent extraction and annotation method of knowledge points for medical textbooks.

[0019] It can be seen from the above technical solutions that the present invention has the following beneficial effects: The embodiment of the present invention provides a method and system for intelligent extraction and annotation of knowledge points for medical textbooks. By collecting and analyzing medical examination questions and textbook resources, using vector representation and semantic similarity evaluation to accurately screen key test points in textbooks, and combining medical concept dictionaries to achieve intelligent annotation, it not only improves the accuracy of information, but also enhances the personalized learning experience. Furthermore, flash cards are designed and generated to promote knowledge memory and verification in the form of intuitive question and answer tables, greatly improving learning efficiency. In addition, the system that integrates intelligent book excerpts, annotations, and flashcard generation functions is seamlessly integrated into the medical students' exam preparation scenarios, reducing manual dependence, and achieving an efficient and convenient exam preparation experience, significantly improving the quality and efficiency of exam preparation compared to traditional methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the implementation cases of the present invention or the technical solutions in the prior art, the following is a brief description of the drawings required for use in the embodiments. By referring to the drawings, the features and advantages of the present invention will be more clearly understood. The drawings are schematic and should not be understood as limiting the present invention in any way. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them: Figure 1 A flowchart of a method for intelligent extraction and annotation of knowledge points for medical textbooks provided in an embodiment; Figure 2 A schematic diagram of the architecture of a knowledge point intelligent extraction and annotation system for medical textbooks provided in an embodiment; Figure 3 A schematic diagram of intelligently extracting key test point excerpts from a textbook in an embodiment; Figure 4 A schematic diagram of intelligently marking key test point excerpts of textbooks in an embodiment; Figure 5 A schematic diagram of generating smart flash cards for textbook key test point book excerpts in an embodiment; Figure 6 It is a schematic diagram of the process of a user using the system of the present invention in an embodiment. DETAILED DESCRIPTION

[0021] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Embodiment 1

[0022] In order to solve the problems of low efficiency, strong dependence on manual work, insufficient personalization, and poor information accuracy in the existing medical learning assistance technology. Figure 1 As shown, the embodiment of the present invention proposes a method for intelligent extraction and annotation of knowledge points for medical textbooks, the method comprising: Step S1: Collect medical examination question resources and medical textbook resources, convert the medical examination question texts and medical textbook texts into vector representations, screen out multiple medical textbook text fragments with the highest semantic similarity to the medical examination question texts by evaluating the semantic similarity between the text fragments, and generate textbook key test point excerpts; Step S2: Based on the constructed medical concept dictionary, the medical concepts in the textbook key test point book excerpts are annotated, and the text fragments related to the real questions in the textbook key test point book excerpts are annotated; Step S3: Design flashcard generation instructions, generate flashcard content in a set format according to the textbook key test point excerpt, and present the flashcard content in a table form containing questions and answers through rendering.

[0023] From the above technical solution, it can be seen that the present invention proposes a method for intelligent extraction and annotation of knowledge points for medical textbooks. By collecting and analyzing medical examination questions and textbook resources, using vector representation and semantic similarity evaluation to accurately screen key test points in textbooks, and combining medical concept dictionaries to achieve intelligent annotation, it not only improves information accuracy, but also enhances personalized learning experience. Furthermore, flash cards are designed and generated to promote knowledge memory and verification in the form of intuitive question and answer tables, greatly improving learning efficiency.

[0024] In step S1, by collecting medical examination question resources and medical textbook resources, the medical examination question text and the medical textbook text are converted into vector representations, and by evaluating the semantic similarity between text fragments, multiple medical textbook text fragments with the highest semantic similarity to the medical examination question text are screened out to generate textbook key test point excerpts, specifically including: Step S11: Collect medical examination question resources and medical textbook resources; Step S12: using a text vectorization model to convert medical examination question text and medical textbook text into vector representation; Step S13: storing the vectorized medical textbook text into a vector database; Step S14: For each real question, a hybrid search strategy of vectorized search and keyword search is performed to screen out multiple medical textbook text fragments with the highest semantic similarity to the medical examination real question text as candidate book excerpt fragments by evaluating the semantic similarity between the text fragments; Step S15: Expert review is performed on the selected candidate excerpt segments to generate a final version of the textbook key examination point excerpt.

[0025] In the process of vectorizing the text of medical textbooks, texts whose length exceeds the upper limit of the text vectorization model are segmented before being vectorized.

[0026] Specifically, the present invention uses artificial intelligence technology, combined with medical examination question resources and medical textbook resources, to automatically generate excerpts of key examination points in medical textbooks, aiming to optimize the scenario for medical students to learn medical knowledge efficiently. Traditional methods rely on the personal experience of medical experts or students, manually reading and marking key knowledge points in textbooks. This process is time-consuming and highly subjective, and can easily lead to a mismatch between learning focus and examination content.

[0027] To solve the above problems, the present invention introduces intelligent book excerpt technology, which reversely traces the core test points in medical textbooks based on medical examination real questions resources. Its core logic is that the medical examination real questions directly reflect the key knowledge points in the textbooks. Through intelligent analysis of the knowledge points examined by the real questions of previous years, and automatically extracting the corresponding content from the textbooks, a textbook key test point book excerpt based on the real question big data is formed. At the technical implementation level, the intelligent book excerpt technology relies on advanced semantic matching technology, that is, evaluating the semantic similarity between text fragments. In the specific implementation process, a hybrid strategy combining text vector matching and key medical concept matching is adopted. This method significantly improves the generation efficiency and accuracy of textbook key test point book excerpts, effectively reduces the learning burden of medical students, enables them to focus on core knowledge points more quickly, and provides strong support for exam preparation.

[0028] As an optional specific implementation method, first collect medical examination question resources and medical textbook resources. Then, use the text-embedding-ada-002 model (text vectorization model) to convert the medical examination question text and the medical textbook text into a high-dimensional vector representation to quantify the semantic relationship between the texts. For long texts that exceed the processing capacity of the model, a segmentation strategy is adopted to ensure the accuracy of the vectorization processing. Subsequently, the vectorized medical textbook text is stored in the Elasticsearch vector database, and the Elasticsearch vector search engine is used to perform an efficient hybrid retrieval strategy of vectorization retrieval and keyword retrieval. By evaluating the semantic similarity between text fragments, multiple medical textbook text fragments with the highest semantic similarity to the medical examination question text are selected as candidate book excerpt fragments. Finally, the selected candidate book excerpt fragments are reviewed by experts to generate the final version of the textbook key test point book excerpt to ensure the accuracy and authority of the book excerpt fragment.

[0029] In step S2, based on the constructed medical concept dictionary, the medical concepts in the textbook key test point book excerpts are annotated, and the text fragments related to the real questions in the textbook key test point book excerpts are annotated, specifically including: Based on the constructed medical concept dictionary, scan the medical concepts in the textbook key points book excerpts, annotate the important medical concepts in the textbook key points book excerpts, and specially annotate the medical concepts that appear in the medical examination test texts; When scanning text fragments in the textbook key test point excerpts, if a text fragment is found to have related real questions, the front-end technology is used to insert a real question mark at the end of the text fragment and provide a link to the question page.

[0030] The construction method of the medical concept dictionary is: by sorting out authoritative medical terminology resources and medical textbook vocabularies, a medical concept dictionary is obtained. The authoritative medical terminology resources include the International Classification of Diseases codes, the Human Phenotype Ontology, and common clinical medical terms released by the National Health Commission; the medical textbook vocabularies include terms such as internal medicine and surgery.

[0031] Specifically, after obtaining the textbook excerpts of key test points based on the big data analysis of past exam questions, the key link to improve the learning efficiency of medical students has turned to the effective reading and memorization of these knowledge fragments. To this end, the present invention introduces intelligent annotation technology, aiming to optimize the learning process through automated means.

[0032] The primary function of intelligent annotation technology is to accurately identify and annotate important medical concepts in textbook texts. These medical concepts, as the cornerstone of medical knowledge, include but are not limited to disease names, typical symptoms, key cells and molecules, etc. This process not only guides learners' attention through intuitive highlighting, but also combines word frequency statistics technology to specifically mark those medical concepts that frequently appear in past exam questions, so as to further strengthen the focus of learning and effectively improve the pertinence and efficiency of learning.

[0033] Secondly, the intelligent annotation technology also realizes the close association mark between the textbook content and the relevant real questions. This means that after completing the study of key medical concepts, learners can immediately self-test through the associated real questions, verify the learning results, and realize the instant feedback loop between theory and practice. The realization of this function relies on advanced text scanning technology (such as Flashtext), which can efficiently load the medical concept dictionary containing authoritative medical terms at home and abroad and textbook-specific vocabulary, conduct in-depth analysis of the text, and use front-end technologies such as HTML and JavaScript to realize the highlighting of medical concepts and the association mark of real questions, including adding eye-catching red dots next to specific medical concepts to indicate their appearance in the real questions, and inserting interactive real question links at the end of the text, so that learners can jump directly to the question-solving interface.

[0034] In summary, intelligent annotation technology, through its refined annotation function and efficient real question association mechanism, not only simplifies the process of learners screening and memorizing key knowledge points, but also promotes the deep integration of theoretical knowledge and practical application, thereby significantly improving the efficiency and effectiveness of medical students' preparation for the exam.

[0035] In step S3, a flashcard generation instruction is designed to generate flashcard content in a set format according to the textbook key test point excerpt, and the flashcard content is presented in a table form containing questions and answers through rendering, specifically including: Design flashcard generation instructions, generate flashcard content in markdown format based on the textbook key test points and book excerpts, use front-end technology to render markdown data, and present the flashcard content in a table format containing questions and answers.

[0036] Specifically, in order to further enhance the memory efficiency of AI technology in assisting medical students in reviewing and preparing for exams, the present invention introduces intelligent flash card generation technology, which is specifically optimized for key exam point excerpts in textbooks. The working principle of this technology is briefly described as follows: When the user selects a section of textbook text and triggers the smart flashcard generation function, the system will seamlessly integrate and call the advanced large language model - Wenxin Qianfan large model, using its powerful natural language processing capabilities to intelligently analyze and refine the core knowledge points of the text. This process relies on innovative prompting generation technology (Prompting), which guides the large model through carefully designed instructions to ensure that the generated flashcards are both professional and efficient.

[0037] Specifically, the flashcard generation prompt is carefully constructed as: "I hope you can make flashcards based on the text I provide as a professional flashcard maker. Instructions for making flashcards: - A flashcard contains a question, answer.

[0038] - Keep your flashcards simple and clear, focusing on the most important information without losing sight of the main idea.

[0039] - Make sure the questions are specific and unambiguous.

[0040] - Answers should use clear and concise language.

[0041] - Answers should contain only one key fact / name / concept / term.

[0042] - Use professional medical expressions throughout.

[0043] - It is prohibited to add questions and answers to content not mentioned in the text.

[0044] Please output your card as a markdown table (question / answer) without any additional text. " After the flashcard content is generated, the markdown format data is converted into an intuitive and easy-to-read table format using front-end technologies such as HTML and JavaScript and presented to the user. This table not only facilitates users to quickly browse and review knowledge points, but also allows users to quickly verify the accuracy of the flashcard content by comparing it with the original textbook text (i.e., excerpts of key test points), effectively preventing the generation of false or erroneous information, thus providing users with an efficient and accurate medical knowledge review tool. Embodiment 2

[0045] Based on systematic thinking, the present invention designs and develops an intelligent knowledge point extraction and annotation system for medical textbooks. The system deeply integrates AI technology, aiming to optimize the user's review process and improve learning efficiency through intelligent means. After completing the core functions such as intelligent book excerpt extraction, intelligent content annotation and intelligent flash card generation, these functions are integrated into a set of applications (APPs) that can be run on mobile devices such as mobile phones and tablets using front-end technologies such as HTML and JavaScript. This design aims to seamlessly integrate AI technology with the user's actual reading and exam preparation scenarios, providing users with an efficient and convenient learning solution.

[0046] The system architecture of the present invention is as follows Figure 2 As shown, a layered design is adopted to ensure efficient collaboration and data flow among components.

[0047] Data layer: Built on MySQL database, it contains multiple key data tables, such as key test point book excerpt data tables, past exam questions data tables, and medical concept glossary, etc., providing solid data support for the system.

[0048] Algorithm layer: Integrates core algorithm modules such as intelligent book excerpt extraction, intelligent content annotation, and intelligent flash card generation, uses advanced artificial intelligence technology to process text data, extract key information, and generate high-quality learning resources.

[0049] The intelligent book excerpt extraction algorithm module is used to collect medical examination question resources and medical textbook resources, convert the medical examination question text and medical textbook text into vector representation, and screen out multiple medical textbook text fragments with the highest semantic similarity to the medical examination question text by evaluating the semantic similarity between text fragments, and generate textbook excerpts of key test points; Intelligent book excerpt annotation algorithm module, which is used to annotate the medical concepts in the textbook key test point book excerpts based on the constructed medical concept dictionary, and annotate the text fragments related to the real questions in the textbook key test point book excerpts; The intelligent flashcard generation algorithm module is used to design flashcard generation instructions, generate flashcard content in a set format according to the textbook key test points and book excerpts, and present the flashcard content in a table form containing questions and answers through rendering.

[0050] Business layer: Use front-end technologies such as HTML and JavaScript to build a user-friendly interactive interface and present complex algorithm processing results to users in an intuitive and easy-to-use manner.

[0051] User operation process: Selecting a textbook: The user first selects the medical textbook to be reviewed through the APP.

[0052] Reading key excerpts: Based on the user's selection, the system displays textbook excerpts of key test points based on intelligent excerpt technology and expert review and confirmation. This step uses intelligent excerpt technology to automatically extract and display the core content of the textbook, helping users quickly grasp the key points of learning and improve reading efficiency.

[0053] Smart annotation and flashcard generation: After the user selects a specific excerpt, the system automatically starts the smart annotation function, highlights the important medical concepts in the excerpt, and associates them with relevant real questions. At the same time, users can click the "Generate Flashcard" button to instantly obtain the smart flashcard corresponding to the excerpt. Flashcards are presented in the form of questions and answers, helping users consolidate their memory and test their knowledge points by practicing questions, further improving learning efficiency.

[0054] Through systematic design and implementation, the entire system achieves a deep integration of AI technology and medical learning scenarios, providing users with an efficient and intelligent solution for reading and preparing for exams.

[0055] In order to further illustrate the technical solution of the present invention, a specific case is described below.

[0056] like Figure 3 As shown in the figure, intelligent extraction of key points in textbook excerpts includes the following steps: Step 1: Acquiring and analyzing real questions. First, accurately select a real question from the question bank of the comprehensive western medicine examination and extract its text content, for example: "Which of the following concepts about the human body's internal environment is incorrect: A. All the liquids in the body B. The environment in which cells survive C. The basic physical and chemical properties change very little D. The result of biological evolution". This step ensures the accuracy and representativeness of the basic data for subsequent analysis.

[0057] Step 2: Text vectorization. Using the text vectorization model text-embedding-ada-002, the above test text is converted into a high-dimensional vector (specifically 1536 dimensions). This vectorization process can capture the semantic features in the text and lay the foundation for the subsequent similarity calculation in the vector space.

[0058] Step 3: Vector-based information retrieval. The converted real-question text vector is used as the query input, and the Elasticsearch vector search engine is used to perform a similarity search in the pre-built and stored textbook text fragment vector database. This process aims to quickly screen out textbook fragments that are closely related to the real-question content and may contain the correct answer or important knowledge points, that is, potential real-question-related book excerpt fragments.

[0059] Step 4: Expert evaluation and optimization. Medical experts will carefully review and evaluate the potential book excerpts retrieved by the computer to confirm their accuracy and teaching value. For those that are not accurate enough or have redundant information, experts will make necessary modifications and optimizations to ensure that the content presented to students is accurate, concise, and helpful for understanding and memory.

[0060] Step 5: Content organization and front-end display. According to the chapter structure and examination syllabus of medical textbooks, the excerpts of key examination points confirmed and optimized by experts are organized and summarized in an orderly manner. Subsequently, using front-end technologies such as HTML and JavaScript, a user-friendly interface is designed and implemented to clearly present key information such as textbook chapters, relevant excerpts of key examination points in each chapter, specific page numbers of excerpts in textbooks, and the frequency of examinations in previous years to users. This process not only improves the pertinence and efficiency of learning, but also helps students establish a direct connection between knowledge and examination questions.

[0061] Intelligent annotation of textbook excerpts of key test points aims to significantly improve users' reading efficiency and learning outcomes. Figure 4 A specific implementation case of intelligent annotation technology is demonstrated: Given a book excerpt with the content "The fluid in the human body is generally called... and does not belong to the category of internal environment", when the user activates the browsing function of the book excerpt, the system immediately starts the automatic annotation mechanism. This mechanism first identifies and highlights the important medical concepts involved in the book excerpt, such as key terms such as "body fluid", "intracellular fluid" and "extracellular fluid". These terms are key test points in the medical knowledge system, and help users quickly focus on key information through visual reinforcement.

[0062] Furthermore, the system also has the ability to analyze the association with historical real questions. If it is detected that a medical concept has appeared in past exam questions, the system will automatically add a red mark point (small red dot) to the upper right corner of the term to intuitively prompt the user the importance of the concept and its frequency of appearance in the exam. In addition, for paragraphs or knowledge points in the book excerpts that are directly related to the real questions of previous years, the system adopts a more obvious annotation method-embedding a logo with a circle and the word "question" in the relevant text area. This design not only indicates the special importance of the content here, but also allows users to jump directly to the relevant real question practice page by clicking on the logo. This function realizes the instant switching between learning and practice, and promotes users to instantly test their mastery of knowledge points during the reading process, thereby improving the effectiveness and pertinence of learning.

[0063] Intelligently extracting and organizing the core knowledge points in large sections of textbook texts plays a significant role in optimizing the user's knowledge system construction, enhancing the logical connection between knowledge points, and improving memory efficiency. Figure 5 A specific implementation case of smart flash card generation technology is demonstrated.

[0064] Given a textbook text "The general name of the fluid in the human body is... It does not belong to the category of the internal environment", the system seamlessly integrates the Wenxin Qianfan model and automatically executes the smart flashcard generation process. This process deeply analyzes the text content, converts it into a series of structured question-answer pairs, and presents it in the form of a smart flashcard in a tabular form. Each flashcard contains a carefully designed "question" and a corresponding "answer" column. For example, the "question" column asks "What is body fluid?", and the "answer" column accurately explains "The general name of the fluid in the human body accounts for about 60% of body weight". This smart flashcard technology not only effectively extracts the key knowledge points in the textbook text, but also strengthens the logical relationship between the knowledge points through the form of questions and answers, so that users can quickly grasp the knowledge context and deepen their understanding when browsing the flashcards. At the same time, breaking down complex knowledge points into concise and clear question-answer pairs not only reduces the difficulty of learning, but also significantly improves the user's learning efficiency and memory consolidation effect through repeated review and testing. Therefore, the smart flashcard generation technology provides users with an efficient and convenient learning tool, which helps them get twice the result with half the effort in the learning process.

[0065] The present invention has finally successfully developed an intelligent extraction and annotation system that integrates AI-assisted content extraction, memory reinforcement and knowledge verification. The system follows a preset, highly structured business logic process, which greatly simplifies the process of users grasping key points, deepening memory and consolidating knowledge mastery through exercises when learning subject knowledge such as physiology. The following is a detailed explanation using the example of a user reviewing physiology. Figure 6 shown.

[0066] After the user enters the computer system, he first clicks on the link of the physiology textbook, and then he can see the key test point book excerpts organized by chapter and syllabus, which are summarized based on artificial intelligence technology and expert confirmation; after clicking, he can enter the book excerpt list page and start studying and reading the key test point book excerpts; by clicking Figure 6 Take the second book excerpt in the book "The fluids in the human body are collectively called body fluids..." as an example. After entering, you can see the page automatically annotated by AI, including the annotations of important concepts and related real questions; at the same time, you can click "Flash Card Generation" to view the smart flash cards containing questions and answers produced by AI; after reviewing the key test point book excerpt, you can click the related real questions and practice to test the mastery of knowledge points in real time. Through the above path, compared with the method that simply relies on traditional manual reading and existing technical solutions, the quality and efficiency of users' reading, remembering key points, and doing questions to test knowledge points can be greatly improved. Embodiment 3

[0067] An embodiment of the present invention provides an electronic device, which includes a processor, a memory and a bus system. The processor and the memory are connected through the bus system. The memory is used to store instructions, and the processor is used to execute the instructions stored in the memory to implement the above-mentioned intelligent extraction and annotation method of knowledge points for medical textbooks. Embodiment 4

[0068] An embodiment of the present invention provides a computer storage medium, which stores a computer software product. The computer software product includes several instructions for enabling a computer device to execute the above-mentioned intelligent extraction and annotation method of knowledge points for medical textbooks.

[0069] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0070] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0071] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide for implementing the process in the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0072] Obviously, the above embodiments are merely examples for clear explanation and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived from them are still within the protection scope of the invention.

Claims

1. A method for intelligent extraction and annotation of knowledge points in medical textbooks. It is characterized in that include: Collect medical exam question resources and medical textbook resources, convert the medical exam question texts and medical textbook texts into vector representations, evaluate the semantic similarity between text fragments, select multiple medical textbook text fragments with the highest semantic similarity to the medical exam question texts, and generate textbook excerpts of key test points; Based on the constructed medical concept dictionary, the medical concepts in the textbook key test point book excerpts are annotated, and the text fragments related to the real questions in the textbook key test point book excerpts are annotated; Design flashcard generation instructions, generate flashcard content in a set format based on the textbook key test points and book excerpts, and present the flashcard content in a table form containing questions and answers through rendering.

2. The method for intelligent extraction and annotation of knowledge points for medical textbooks according to claim 1, It is characterized in that Collect medical exam question resources and medical textbook resources, convert the medical exam question text and medical textbook text into vector representation, evaluate the semantic similarity between text fragments, select multiple medical textbook text fragments with the highest semantic similarity to the medical exam question text, and generate textbook key points excerpts, including: Step S11: Collect medical examination question resources and medical textbook resources; Step S12: using a text vectorization model to convert medical examination question text and medical textbook text into vector representation; Step S13: storing the vectorized medical textbook text into a vector database; Step S14: For each real question, a hybrid search strategy of vectorized search and keyword search is performed to screen out multiple medical textbook text fragments with the highest semantic similarity to the medical examination real question text as candidate book excerpt fragments by evaluating the semantic similarity between the text fragments; Step S15: Expert review is performed on the selected candidate excerpt segments to generate a final version of the textbook key examination point excerpt.

3. The method for intelligent extraction and annotation of knowledge points for medical textbooks according to claim 2, It is characterized in that In the process of vectorizing the text of medical textbooks, texts whose length exceeds the upper limit of the text vectorization model are segmented before being vectorized.

4. The method for intelligent extraction and annotation of knowledge points for medical textbooks according to claim 1, It is characterized in that The method for constructing the medical concept dictionary is: A medical concept dictionary is obtained by collating authoritative medical terminology resources and medical textbook vocabularies, wherein the authoritative medical terminology resources include the International Classification of Diseases codes, the Human Phenotype Ontology, and commonly used clinical medical terms released by the National Health Commission; the medical textbook vocabularies include internal medicine and surgery.

5. The method for intelligent extraction and annotation of knowledge points for medical textbooks according to claim 1, It is characterized in that Based on the constructed medical concept dictionary, the medical concepts in the textbook key test point book excerpts are annotated, and the text fragments related to the real questions in the textbook key test point book excerpts are annotated, including: Based on the constructed medical concept dictionary, scan the medical concepts in the textbook key points book excerpts, annotate the important medical concepts in the textbook key points book excerpts, and specially annotate the medical concepts that appear in the medical examination test texts; When scanning text fragments in the textbook key test point excerpts, if a text fragment is found to have related real questions, the front-end technology is used to insert a real question mark at the end of the text fragment and provide a link to the question page.

6. The method for intelligent extraction and annotation of knowledge points for medical textbooks according to claim 1, It is characterized in that Design flashcard generation instructions, generate flashcard content in a set format based on the textbook key points and excerpts, and render the flashcard content into a table containing questions and answers, including: Design flashcard generation instructions, generate flashcard content in markdown format based on the textbook key test points and book excerpts, use front-end technology to render markdown data, and present the flashcard content in a table format containing questions and answers.

7. The method for intelligent extraction and annotation of knowledge points for medical textbooks according to claim 5 or 6, It is characterized in that The front-end technologies include HTML and JavaScript.

8. An intelligent knowledge extraction and annotation system for medical textbooks. It is characterized in that The system is used to implement the intelligent extraction and annotation method of knowledge points for medical textbooks as described in any one of claims 1 to 7, specifically comprising: Intelligent book excerpt extraction algorithm module, used to collect medical examination question resources and medical textbook resources, convert medical examination question text and medical textbook text into vector representation, and select multiple medical textbook text fragments with the highest semantic similarity to medical examination question text by evaluating the semantic similarity between text fragments, and generate textbook excerpts of key examination points; Intelligent book excerpt annotation algorithm module, which is used to annotate the medical concepts in the textbook key test point book excerpts based on the constructed medical concept dictionary, and annotate the text fragments related to the real questions in the textbook key test point book excerpts; The intelligent flashcard generation algorithm module is used to design flashcard generation instructions, generate flashcard content in a set format according to the textbook key test points and book excerpts, and present the flashcard content in a table form containing questions and answers through rendering.

9. An electronic device, It is characterized in that The electronic device includes a processor, a memory and a bus system, the processor and the memory are connected through the bus system, the memory is used to store instructions, and the processor is used to execute the instructions stored in the memory to implement the intelligent extraction and annotation method of knowledge points for medical textbooks as described in any one of claims 1 to 7.

10. A computer storage medium, It is characterized in that The computer storage medium stores a computer software product, and the computer software product includes several instructions for enabling a computer device to execute the intelligent extraction and annotation method of knowledge points for medical textbooks as described in any one of claims 1 to 7.

Citation Information

Patent Citations

  • Data interaction method, device and storage medium

    CN112905768A

  • Question and answer method and device, electronic equipment and storage medium

    CN117708290A

Cited By

  • An intelligent annotation method and system for semantic structural units in medical texts

    CN119783681A