Automatic medicine research and development electronic newspaper pushing method and device based on large model

Abstract

The invention discloses an automatic medicine research and development electronic newspaper pushing method and device based on a large model, and the device comprises a user interaction module which is used for providing an information subscription function, an information browsing function and a consultation question and answer function which can be interacted for a user; the data acquisition module is used for acquiring medicine research and development related data from different data sources; the large model processing module is used for intelligently analyzing and processing the acquired medicine research and development related data through a large model; the intelligent pushing module is used for performing content screening and matching on the information subjected to large model processing according to subscription preferences of the user, performing automatic typesetting to form electronic newspaper, displaying the electronic newspaper and pushing the electronic newspaper to the user; and the data storage module is used for storing various data, including processed information and user data. According to the device and the method, comprehensive, timely and accurate information service is provided for medicine research and development personnel through multi-source data acquisition, intelligent processing analysis and personalized pushing.

Description

Technical Field

[0001] The present invention belongs to the technical field of information processing and intelligent push, and specifically relates to a method and device for pushing an automated medical research and development electronic newspaper based on a large model. Background Art

[0002] In the field of pharmaceutical research and development, drug developers pay close attention to the update of first-hand information such as scientific research literature, clinical trial results, academic conference content, and pharmaceutical company research and development dynamics. However, existing information acquisition tools are mainly based on digital document retrieval and abstract generation. Although this model is convenient, the information presentation lacks integration and intuitiveness. Users often need to obtain scattered information on multiple platforms, resulting in content fragmentation, which further reduces research and development efficiency.

[0003] Current scientific research assistant tools, such as the biomedical Yazai launched by Zhihuiya and the AI ​​intelligent and efficient reading tool scholarcy, mostly rely on users to upload PDF documents and cannot push updated content on a journal basis. More importantly, these tools find it difficult to organically integrate multi-source data updated in real time, and cannot meet users' needs to quickly and comprehensively understand the latest developments. Summary of the invention

[0004] In view of the technical problems that existing scientific research assistants based on big models are heavily dependent on users uploading PDFs, information is fragmented, and the presentation format is single, the purpose of the present invention is to provide an automated medical research and development electronic newspaper push method and device based on big models, which provides comprehensive, timely and accurate information services for medical research and development personnel through multi-source data collection, intelligent processing and analysis, and personalized push.

[0005] To achieve the above-mentioned purpose of the invention, an embodiment provides an automated medical research and development electronic newspaper push device based on a large model, comprising: The user interaction module is used to provide users with interactive information subscription functions, information browsing functions, and consultation and question-and-answer functions; Data collection module, which is used to collect data related to pharmaceutical research and development from different data sources; A large model processing module, which is used to intelligently analyze and process the collected medical research and development related data through a large model; Intelligent push module, which is used to automatically layout the information processed by the large model into an electronic newspaper according to the user's subscription preferences, and push it to the user; And a data storage module, which is used to store various types of data, including processed information and user data.

[0006] Preferably, the user interaction module includes: The information subscription interface provides the information subscription function. Through the information subscription function, users can search and filter the information source types according to their needs. They can also customize the combination of information source types and adjust the priority of certain information sources. Among them, the information source types include journal articles, public account articles, clinical trial data, academic conference information, and pharmaceutical company news; The information browsing interface provides information browsing functions. Users can use the information browsing function to filter information based on the filtering conditions and display the key information processed by the large model. The filtering conditions include keywords, topics, and release dates. As well as an information question and answer interface, it provides an information question and answer function, through which users can ask in-depth questions about the content of interest, and supports saving the answer results to personal favorites.

[0007] Preferably, the data acquisition module includes: Journal data collection function, which is used to monitor journal updates online in real time and fetch the latest published medical papers from the corresponding journal webpage in real time; Public account article collection function, which is used to collect the latest medical-related information from the public accounts subscribed by users; Clinical trial data collection function, which is used to collect the latest clinical trial information from major clinical trial databases; Academic conference information collection function, which is used to obtain cutting-edge dynamics and academic exchange content from medical-related academic conferences; As well as the pharmaceutical company dynamics collection function, which is used to track the latest R&D progress, market trends and major news of pharmaceutical companies.

[0008] Preferably, the large model processing module includes: Information transmission channel, which is used to provide a high-speed and secure data transmission interface for large models; The dynamic model adjustment function is used to dynamically select the corresponding large model pipeline according to the task type and design prompt words suitable for large model understanding; The big model generation function is used to generate prompt text based on the medical research and development related data combined with prompt words, and call the big model related to the task type to perform intelligent analysis and processing based on the prompt text to obtain processed information.

[0009] Preferably, the intelligent push module includes: Content screening and matching function, which is used to screen and match the information processed by the big model with the user's reading preferences; Automatic layout and display function, which is used to automatically layout the screening results and matching content to generate electronic newspapers, where the electronic newspapers are presented in a layout with special classification, summary aggregation, and key points highlighted, and support PDF download, electronic reading and printing; The intelligent push function is used to push the electronic newspapers corresponding to the most relevant matching content in priority according to the matching degree. The push forms include web page display and email push.

[0010] Preferably, the content screening and matching function is implemented through the following three agents: The manuscript screening agent collects the pharmaceutical R&D-related data through the data collection module, combines the keyword matching algorithm, topic tracking function and user interest model, and sorts and screens the pharmaceutical R&D-related data. The screening results ensure the relevance of the news content to the user's subscribed topics, and store them in the data storage module in the form of structured data, and the data is sorted by date; The content creation agency uses the big model module to generate multiple versions of news articles that meet user needs, covering both in-depth reports and short news; The proofreading and optimization agent will call the large model module to automatically detect and correct grammatical errors, logical loopholes and content redundancy in news releases.

[0011] Preferably, the automatic typesetting and display function is realized through the following process: The typesetting layout algorithm generates an electronic newspaper based on the proofread and optimized news articles and corresponding illustrations. The typesetting layout algorithm dynamically adjusts the layout of the electronic newspaper through the following rules: automatically selects a matching template based on the number of words and pictures in the article; optimizes the ratio of text to pictures; and automatically generates content-related illustrations for articles without illustrations.

[0012] Preferably, the data storage module comprises: The processing result storage function is used to save only the information processed by the large model, including classification, labeling, summary and key information after knowledge extraction; User data storage function, which is used to record user subscription settings, interest preferences, usage data, and is also used to generate user portraits and personalized push strategies. It also stores user collections and historical records.

[0013] To achieve the above-mentioned purpose of the invention, the embodiment of the present invention further provides a method for pushing an automated medical research and development electronic newspaper based on a large model, the method using the above-mentioned automated medical research and development electronic newspaper pushing device, the method comprising the following steps: Collect medical research and development related data from different data sources through the data collection module; Interact through the information subscription function, information browsing function, and consultation and question-answering function provided by the user interaction module, set subscription preferences, browse information, and conduct consultation and question-answering; Through the big model processing module, the collected medical research and development related data are intelligently analyzed and processed through the big model; Through the intelligent push module, according to the user's subscription preferences, the information processed by the large model is screened and matched, and then automatically typeset to form an electronic newspaper display and push it to the user; Various types of data are stored through the data storage module, including processed information and user data.

[0014] To achieve the above-mentioned purpose of the invention, an embodiment of the present invention further provides a computer-readable storage medium on which a program is stored. When the program is executed by a processor, the above-mentioned large-model-based automated medical research and development electronic newspaper push method is implemented.

[0015] Compared with the prior art, the present invention has the following beneficial effects: The present invention innovatively displays updated information content by journal, generates electronic newspapers by integrating the core documents of each issue of the journal, and provides thematic structured content, so that users can quickly understand the latest research trends; Through in-depth analysis and processing of multi-source data, the present invention can orderly integrate and correlate information scattered across different channels, generate clearly structured thematic content and concise and clear electronic newspaper-style output, help users quickly understand key research information from multiple dimensions, and solve the problems of isolated and fragmented information in existing tools. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0017] Figure 1 It is a structural schematic diagram of an automated medical research and development electronic newspaper push device based on a large model provided in an embodiment; Figure 2 It is a flow chart of the method for pushing electronic newspapers of automated medical research and development based on a large model provided in an embodiment. DETAILED DESCRIPTION

[0018] To make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific implementation methods described herein are only used to explain the present invention and do not limit the scope of protection of the present invention.

[0019] The inventive concept of the present invention is: in order to solve the technical problems that the existing scientific research assistant based on the big model is heavily dependent on the user to upload PDF, the information is fragmented, and the display form is single, the embodiment of the present invention provides an automated medical research and development electronic newspaper push solution based on the big model, focusing on converting the updated content in the field of medical research and development into an intelligent newspaper form, and integrating and pushing the latest dynamics to users in a manner similar to traditional newspapers through column division, classification and summary processing. At the same time, an innovative information push method with journals as units is proposed to meet the needs of multiple scenarios. The method uses the big model technology to perform online parsing and in-depth analysis of multi-source data, and combines the automatic typesetting technology to present the important updated content in various fields in a clear and intuitive layout form. Users can obtain content through paper printing, PDF electronic newspapers or web browsing, etc., to meet the use needs of multiple scenarios. This innovative newspaper output method not only solves the problem of information fragmentation, but also provides a form of information display that is easy to read quickly and highlights the key points, providing a new information acquisition experience for medical research and development personnel, and effectively improving the efficiency of information management and application.

[0020] like Figure 1 As shown, the automated medical research and development electronic newspaper push device provided in the embodiment includes a user interaction module 110, an intelligent push module 120, a data acquisition module 130, a data storage module 140, and a large model processing module 150. These modules form a technical solution for multi-source data acquisition, intelligent processing and personalized push, ensuring that medical practitioners can obtain first-hand information efficiently and conveniently. The large model here refers to an artificial intelligence model with a large scale and complex structure. They have hundreds of millions of parameters and a deep neural network architecture. By learning massive data and optimizing deep neural networks, they have achieved remarkable results in various tasks.

[0021] In the embodiment, the user interaction module 110 is used to provide users with interactive information subscription functions, information browsing functions, and consultation and question-answering functions. Specifically, the user interaction module includes an information subscription interface, an information browsing interface, and an information question-answering interface.

[0022] The information subscription interface provides an information subscription function, through which users can search and filter information source types according to their needs, and can also customize the combination of information source types and adjust the priority of certain information sources in order to obtain first-hand information that meets their own information needs. Among them, information source types include journal articles, public account articles, clinical trial data, academic conference information, and pharmaceutical company news.

[0023] The information browsing interface provides a friendly and easy-to-use interface for browsing, content retrieval, and collection of information. The information browsing interface is the main window for users to obtain information, and the pushed content is organized and displayed according to user preferences. The information browsing interface provides a convenient information browsing function. Users can consult and filter according to the filtering conditions through the information browsing function, quickly find the required content, and clearly display the key information processed by the large model on the interface. The filtering conditions include keywords, topics, publication dates, journal articles, public account articles, clinical trial numbers, academic conferences, and pharmaceutical companies.

[0024] The information question and answer interface provides an information question and answer function, through which users can ask in-depth questions about content of interest, and support saving the answer results to personal favorites for subsequent viewing and management.

[0025] Based on the above user interaction module, users can conduct personalized searches through keywords. Users can set subscription preferences for different types of information based on their research directions and interests on the platform, including specific journals, research topics, clinical trial numbers, or dynamics of specific pharmaceutical companies, etc. This helps users obtain the information they need more accurately.

[0026] The above user interaction module provides a powerful search function. Users can filter and search for medical information on the platform by keywords, research fields, dates, etc. The search results are displayed in card or list format. Each piece of information contains a title, abstract, and key information, which is convenient for users to browse quickly.

[0027] Users can conduct in-depth Q&A on the information content they are interested in, and save the answers to their personal favorites. The device will group and manage the collections according to research directions or projects. Users can customize group collections according to research directions or topics for future search and reference. The collections can be viewed and managed at any time, and the export function is supported, which is convenient for users to cite and refer to in subsequent research.

[0028] In the embodiment, the data collection module 130 is used to collect data related to pharmaceutical research and development from different data sources. In order to ensure the extensiveness of data sources, it specifically includes journal data collection functions, public account article collection functions, clinical trial data collection functions, academic conference information collection functions, and pharmaceutical company dynamics collection functions. In the data collection process, API interfaces and crawler technology are used to achieve real-time updates, and data is captured within preset intervals through scheduled tasks to ensure the timeliness of system content. At the same time, it is also equipped with data cleaning and format conversion functions to convert raw data from different sources into a unified format for further analysis by the large model processing module.

[0029] Among them, the journal data collection function is used to monitor journal updates online in real time, and to capture the latest published medical papers from the corresponding journal webpage in real time. When collecting journal data, the latest medical research papers are regularly captured through API interfaces or crawler technology, and incremental updates are performed, and journal content summaries are also performed. Specifically, the incremental crawler update includes: intelligent incremental crawling mechanism: by maintaining document index records, the incremental crawler can monitor the update of the target journal database regularly or in real time, and automatically identify new documents; wide support for journal publishers: the device of the present invention supports the collection of journal OA and non-OA articles from 20 mainstream academic publishers, including Springer (Springer), Elsevier (Elsevier), Wiley (John Wiley), AACR (American Association for Cancer Research), ACS (American Chemical Society), AAAS (American Association for the Advancement of Science), Taylor & Francis (Taylor & Francis), IEEE (Institute of Electrical and Electronics Engineers), etc., covering more than 95% of all journal documents. Intelligent update control and scheduling: The built-in intelligent scheduling system can dynamically adjust the frequency and concurrency of document crawling according to factors such as server load and network bandwidth to ensure that document crawling does not violate the crawling object request restrictions and efficiently collect data under legal conditions.

[0030] The journal content summary includes: automatically obtaining all the literature content of the updated journal, and using large models to analyze and integrate the literature content, helping users to quickly understand the core content of the journal update and improve the efficiency of journal reading and screening.

[0031] Among them, the public account article collection function is used to collect the latest medical-related information from the public accounts subscribed by users. Specifically, it monitors the WeChat public accounts that users follow in real time, detects the release of new articles, and collects content related to medical research and development into the database. The specific implementation includes public account retrieval, public account monitoring, and content summary. Public account retrieval refers to automatically retrieving and locating qualified WeChat public accounts by entering keywords, so that the target public accounts can be efficiently screened out from a large number of public accounts, providing a basis for subsequent data collection; public account monitoring is to monitor the target public account in real time, automatically detect the update of public account articles, and when new push content is detected, it will be recorded and collected immediately to ensure the timeliness and integrity of the data; content summary refers to intelligent analysis of article content, extracting the core information in the article and generating a concise content summary, so that users can quickly understand the main content of the article and reduce the time dependence on full-text reading.

[0032] Among them, the clinical trial data collection function is used to collect the latest clinical trial information from major clinical trial databases. The specific synchronization includes connecting to major global clinical trial databases to obtain new clinical trial progress, trial phases, and result data to ensure the comprehensiveness and timeliness of the data. The specific implementation includes multi-source database synchronization and status tracking. Multi-source database synchronization refers to support for connecting to multiple clinical trial databases, including clinical trial databases in the United States, Europe, China, and India. Through multi-source database synchronization, clinical trial information on a global scale can be summarized, database content can be synchronized regularly, and information can be unified for easy retrieval and analysis; status tracking refers to the ability to track the real-time progress status of clinical trials collected by users, such as trial recruitment, ongoing, completed, etc., to help users understand the different stages and key nodes of drug research and development.

[0033] Among them, the academic conference information collection function is used to obtain cutting-edge dynamics and academic exchange content from medical-related academic conferences. Specifically, it obtains conference summaries, reports and research results from the official websites of various academic conferences and integrates the latest academic dynamics.

[0034] The pharmaceutical company dynamics collection function is used to track the latest R&D progress, market trends and major news of pharmaceutical companies, and collect these trends for analysis.

[0035] All data collected by the above data collection module are cleaned, including deduplication, format standardization and elimination of irrelevant data, to provide a good foundation for subsequent analysis.

[0036] In the embodiment, the large model processing module 150 is a processing engine for intelligent analysis and deep processing of the collected medical research and development related data through a large model. The large model processing module 150 has dynamic adjustment capabilities, can design suitable large model prompt words according to different task types (such as translation, text summary generation, and image-generated text), and select the corresponding large model interface for parsing. The large model processing module 150 supports a variety of large language models, such as OpenAI, Google, Baidu, etc., and can automatically update to keep the latest version.

[0037] Specifically, the large model processing module 150 includes an information transmission channel, a model dynamic adjustment function, and a large model generation function. The information transmission channel is used to provide a high-speed and secure data transmission interface for the large model, ensuring that the large model can quickly obtain information for analysis.

[0038] The model dynamic adjustment function is used to dynamically select the corresponding large model pipeline according to the task type and design prompt words suitable for large model understanding.

[0039] The big model generation function is used to generate prompt text based on the pharmaceutical research and development related data combined with prompt words, and call the big model related to the task type to perform intelligent analysis and processing based on the prompt text to obtain processed information. Specific intelligent analysis and processing include using the big model to identify key entities in the text (such as drug names, research institutions, test numbers, test data and research results, etc.), automatically identifying entities and relationships, and classifying based on the semantic understanding of the content. By automatically identifying entities and relationships, it can provide users with information summaries; it also has a summary function to generate a concise summary for each document or article to help users quickly understand the core content of the document.

[0040] In the embodiment, the intelligent push module 120 is used to automatically layout and display electronic newspapers to users after the information processed by the large model is screened and matched according to the user's subscription preferences, and then pushes them to users. Specifically, it includes content screening and matching functions, automatic layout and display functions, and intelligent push functions.

[0041] The content screening and matching function is used to screen and match the information processed by the large model with the user's reading preferences. Specifically, the content screening and matching is implemented through the following three agents: The manuscript screening agent collects the pharmaceutical R&D-related data through the data collection module, combines the keyword matching algorithm, topic tracking function and user interest model, and sorts and screens the pharmaceutical R&D-related data. The screening results ensure the relevance of the news content to the user's subscribed topics, and store them in the data storage module in the form of structured data, and the data is sorted by date; The content creation agent uses the large model module to generate multiple versions of news articles that meet user needs, covering two forms: in-depth reports and short news. The specific generation template example is as follows: "You are a professional news writer. Your task is to create a title and multiple versions of news content based on the given text separated by ###. The title must be accurate, concise and eye-catching. The news content includes short news and in-depth reports. The following are the specific requirements for both: Short News: Write a short news release for general readers, which should be concise, clear and attractive; In-depth analysis: Expanded into an 800-word in-depth analysis article, highlighting the background, impact and expert opinions of the event; ### {article} ### Return in json format, the format template is: { “title”: title, “shotNews”: short news, “deeply interpret”: } ” Article is the content of each article.

[0042] The proofreading and optimization agent will call the big model module to automatically detect and correct grammatical errors, logical loopholes and content redundancy in news releases. The specific template generation example is as follows: The following is a news article and its feedback information. Please check whether there are any errors or incoherence in the grammar and semantics, and whether there are any incoherent expression logic. Please first determine whether this article needs to be modified. If no modification is required, please return None directly; if modification is required, modify it according to the following requirements: 1. Correct grammar and sentence structure and improve language fluency while maintaining consistency in content; 2. Add necessary connecting words to make the whole text more coherent; Please provide an optimized version. Manuscript: '{News release}', Feedback: '{Feedback content}'. ” The automatic typesetting and display function is used to automatically typeset based on the screening results and matching content to generate an electronic newspaper. The electronic newspaper is presented in a typesetting manner with thematic classification, abstract aggregation, and highlighted key points. It covers information such as journal update briefs, clinical trial dynamics, academic conference reports, and pharmaceutical company news, allowing users to quickly browse the entire issue and clearly grasp the overall dynamics. It supports PDF downloads, electronic reading, and printing, ensuring that users can quickly and intuitively obtain the latest information while meeting the usage requirements of different scenarios. When automatically typesetting, the typesetting layout algorithm generates an electronic newspaper based on the proofread and optimized news releases and corresponding illustrations. The typesetting layout algorithm dynamically adjusts the layout of the electronic newspaper through the following rules: automatically select a matching template based on the number of words and pictures in the article; optimize the ratio of text to pictures; and automatically generate content-related illustrations for articles without illustrations.

[0043] The smart push function is used to prioritize the e-newspapers corresponding to the most relevant matching content according to the degree of matching. The push forms include web page display and email push, ensuring that users can receive content through multiple channels.

[0044] In the embodiment, the data storage module 140 is used to store various types of data, including processed information and user data. In order to achieve large-scale data management and fast access, the data storage module 140 adopts a distributed storage architecture, which can maintain the stability and response speed of the system under high concurrent requests. Specifically, it includes a processing result storage function and a user data storage function, wherein the processing result storage function is used to not save the original data, but only save the information content processed by the large model, including classification, labeling, summary and key information after knowledge extraction. The storage of processing results provides a basis for push and user retrieval, enabling the system to quickly respond to user requests; the user data storage function is used to record the user's subscription settings, interest preferences, and usage data, and is also used for the generation of user portraits and personalized push strategies. At the same time, it also stores the user's favorite content and historical records to provide data support for user interaction.

[0045] Based on the above-mentioned automated electronic newspaper push device for pharmaceutical research and development, firstly, the data acquisition module is used to obtain pharmaceutical-related information from multiple channels, mainly focusing on the update of pharmaceutical journals, and secondly, it also includes public accounts, clinical trials, academic conference information and pharmaceutical company dynamics that users are interested in; the collected data is classified, extracted, summarized and analyzed using a large model; according to the user's subscribed topics and preferences, the relevant information is pushed to the user in the form of an electronic newspaper through the intelligent push module; the user can interactively explore and collect the pushed content. The device of the present invention mainly aims at the problem that it is cumbersome for pharmaceutical researchers to obtain the latest research dynamics and the information is seriously fragmented. It adopts a large model-based method to integrate and push information, improve the efficiency of researchers in obtaining information, and provide in-depth analysis insights, providing all-round information support for pharmaceutical research and development.

[0046] Based on the above-mentioned automatic medical research and development electronic newspaper pushing device, the embodiment further provides an automatic medical research and development electronic newspaper pushing method based on a large model, comprising the following steps: S1, collects pharmaceutical R&D related data from different data sources through the data collection module; S2, interacting through the information subscription function, information browsing function, and consultation and question-answering function provided by the user interaction module, setting subscription preferences, and conducting information browsing and consultation and question-answering; S3, through the big model processing module, intelligently analyzes and processes the collected medical research and development related data through the big model; S4, through the intelligent push module, according to the user's subscription preferences, the information processed by the large model is screened and matched, and then automatically typeset to form an electronic newspaper display and push it to the user; S5, stores various types of data through the data storage module, including processed information and user data.

[0047] The embodiment further provides a computer-readable storage medium on which a program is stored. When the program is executed by a processor, the above-mentioned method for pushing the automated medical research and development electronic newspaper based on a large model is implemented, which specifically includes the following steps: S1, collects pharmaceutical R&D related data from different data sources through the data collection module; S2, interacting through the information subscription function, information browsing function, and consultation and question-answering function provided by the user interaction module, setting subscription preferences, and conducting information browsing and consultation and question-answering; S3, through the big model processing module, intelligently analyzes and processes the collected medical research and development related data through the big model; S4, through the intelligent push module, according to the user's subscription preferences, the information processed by the large model is screened and matched, and then automatically typeset to form an electronic newspaper display and push it to the user; S5, stores various types of data through the data storage module, including processed information and user data.

[0048] In the embodiment, computer-readable media include permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. The information can be computer-readable instructions, data structures, program modules or other data.

[0049] The specific implementation methods described above provide a detailed description of the technical solutions and beneficial effects of the present invention. It should be understood that the above is only the most preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, supplements and equivalent substitutions made within the scope of the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An automated medical research and development electronic newspaper push device based on a large model, characterized in that: include: The user interaction module is used to provide users with interactive information subscription functions, information browsing functions, and consultation and question-and-answer functions; Data collection module, which is used to collect data related to pharmaceutical research and development from different data sources; A large model processing module, which is used to intelligently analyze and process the collected medical research and development related data through a large model; Intelligent push module, which is used to automatically layout the information processed by the large model into an electronic newspaper according to the user's subscription preferences, and push it to the user; and a data storage module, which is used to store various types of data, including processed information and user data.

2. The large-model-based automated medical research and development electronic newspaper push device according to claim 1, characterized in that: The user interaction module comprises: The information subscription interface provides the information subscription function. Through the information subscription function, users can search and filter the information source types according to their needs. They can also customize the combination of information source types and adjust the priority of certain information sources. Among them, the information source types include journal articles, public account articles, clinical trial data, academic conference information, and pharmaceutical company news; The information browsing interface provides information browsing functions. Users can use the information browsing function to filter information based on the filtering conditions and display the key information processed by the large model. The filtering conditions include keywords, topics, and release dates. As well as an information question and answer interface, it provides an information question and answer function, through which users can ask in-depth questions about content of interest, and supports saving the answer results to personal favorites.

3. The large-model-based automated medical research and development electronic newspaper push device according to claim 1, characterized in that: The data acquisition module comprises: Journal data collection function, which is used to monitor journal updates online in real time and fetch the latest published medical papers from the corresponding journal webpage in real time; Public account article collection function, which is used to collect the latest medical-related information from the public accounts subscribed by users; Clinical trial data collection function, which is used to collect the latest clinical trial information from major clinical trial databases; Academic conference information collection function, which is used to obtain cutting-edge dynamics and academic exchange content from medical-related academic conferences; As well as the pharmaceutical company dynamics collection function, which is used to track the latest R&D progress, market trends and major news of pharmaceutical companies.

4. The large-model-based automated medical research and development electronic newspaper push device according to claim 1, characterized in that: The large model processing module includes: Information transmission channel, which is used to provide a high-speed and secure data transmission interface for large models; The dynamic model adjustment function is used to dynamically select the corresponding large model pipeline according to the task type and design prompt words suitable for large model understanding; The big model generation function is used to generate prompt text based on the medical research and development related data combined with prompt words, and call the big model related to the task type to perform intelligent analysis and processing based on the prompt text to obtain processed information.

5. The large-model-based automated medical research and development electronic newspaper push device according to claim 1, characterized in that: The intelligent push module includes: Content screening and matching function, which is used to screen and match the information processed by the big model with the user's reading preferences; Automatic layout and display function, which is used to automatically layout the screening results and matching content to generate electronic newspapers, where the electronic newspapers are presented in a layout with special classification, summary aggregation, and key points highlighted, and support PDF download, electronic reading and printing; The intelligent push function is used to push the electronic newspapers corresponding to the most relevant matching content in priority according to the matching degree. The push forms include web page display and email push.

6. The large-model-based automated medical research and development electronic newspaper push device according to claim 5, characterized in that: The content screening and matching function is implemented through the following three agents: The manuscript screening agent collects the pharmaceutical R&D-related data through the data collection module, combines the keyword matching algorithm, topic tracking function and user interest model, and sorts and screens the pharmaceutical R&D-related data. The screening results ensure the relevance of the news content to the user's subscribed topics, and store them in the data storage module in the form of structured data, and the data is sorted by date; The content creation agency uses the big model module to generate multiple versions of news articles that meet user needs, covering both in-depth reports and short news; The proofreading and optimization agent will call the large model module to automatically detect and correct grammatical errors, logical loopholes and content redundancy in news releases.

7. The large-model-based automated medical research and development electronic newspaper push device according to claim 6, characterized in that: The automatic layout display function is realized through the following process: The typesetting layout algorithm generates an electronic newspaper based on the proofread and optimized news articles and corresponding illustrations. The typesetting layout algorithm dynamically adjusts the layout of the electronic newspaper through the following rules: automatically selects a matching template based on the number of words and pictures in the article; optimizes the ratio of text to pictures; and automatically generates content-related illustrations for articles without illustrations.

8. The large-model-based automated medical research and development electronic newspaper push device according to claim 1, characterized in that: The data storage module comprises: The processing result storage function is used to save only the information processed by the large model, including classification, labeling, summary and key information after knowledge extraction; User data storage function, which is used to record user subscription settings, interest preferences, usage data, and is also used to generate user portraits and personalized push strategies. It also stores user collections and historical records.

9. A method for pushing electronic newspapers of automated medical research and development based on a large model, characterized in that: The method uses the automated medical research and development electronic newspaper push device according to any one of claims 1 to 8, and the method comprises the following steps: Collect medical research and development related data from different data sources through the data collection module; Interact through the information subscription function, information browsing function, and consultation and question-answering function provided by the user interaction module, set subscription preferences, browse information, and conduct consultation and question-answering; Through the big model processing module, the collected medical research and development related data are intelligently analyzed and processed through the big model; Through the intelligent push module, according to the user's subscription preferences, the information processed by the large model is screened and matched, and then automatically typeset to form an electronic newspaper display and push it to the user; Various types of data are stored through the data storage module, including processed information and user data.

10. A computer-readable storage medium, characterized in that: A program is stored thereon, and when the program is executed by the processor, the method for pushing the automated medical research and development electronic newspaper based on a large model as described in claim 9 is implemented.

Images