Intelligent information analysis system and method based on multi-modal technology subject track

Abstract

The application relates to an intelligent intelligence analysis system and method based on a multi-modal technology theme track, wherein the system comprises: a homepage module for establishing an index of a patent database and / or a paper database to retrieve patent data and / or paper data; a multi-modal data search module for performing semantic retrieval on a target database by using a preset model context protocol to obtain retrieval results; an intelligent analysis module for cleaning and standardizing the retrieval results to obtain final data, and identifying intelligence generation requirements of a user according to an analysis type selected by the user; an AI-driven engine for generating a corresponding professional analysis report according to the final data and the intelligence generation requirements by using a preset large language model; and a visual display module for generating multi-dimensional data visualization information based on the professional analysis report to display the professional analysis report to the user. Thus, the problems in the prior art that only a single data source is supported, multi-modal data comprehensive analysis cannot be integrated, retrieval efficiency is low and response time is long, real-time interaction requirements cannot be met, deep analysis is difficult, and intelligent analysis report generation capability is lacking are solved.

Description

Technical Field

[0001] This application relates to the fields of science and technology intelligence analysis and knowledge graph technology, and in particular to an intelligent intelligence analysis system and method based on multimodal science and technology topic trajectories. Background Technology

[0002] With the globalization of technological innovation and increasingly fierce competition, the importance of intelligent intelligence analysis of technology-related trajectories for enterprise technological innovation and national strategic planning has significantly increased. Traditional technology intelligence analysis mainly relies on manual retrieval and analysis, which suffers from low efficiency, narrow coverage, and insufficient analytical depth. With the development of big data and artificial intelligence technologies, automated technology intelligence analysis systems are gradually emerging.

[0003] In related technologies, science and technology intelligence analysis tools such as Derwent Innovation and PatSnap provide basic patent analysis functions, supporting retrieval, statistics and visualization of single data sources (such as patents only or papers only). Derwent Innovation focuses on patent data processing, realizing patent classification number retrieval and applicant technology layout statistics. PatSnap supplements the patent data with simple technology efficacy matrix analysis.

[0004] However, the relevant technologies only support a single data source, cannot integrate multimodal data for comprehensive analysis, have low retrieval efficiency, and when faced with hundreds of millions of data points, the response time often cannot meet the needs of real-time interaction. They are difficult to conduct in-depth trend analysis, collaborative analysis, and comparative analysis, lack intelligent analysis report generation capabilities, and cannot automatically produce intelligence content with professional insights. These issues urgently need improvement. Summary of the Invention

[0005] This application provides an intelligent intelligence analysis system and method based on multimodal technology theme trajectories to solve the problems in related technologies, such as only supporting a single data source, resulting in the inability to integrate multimodal data for comprehensive analysis, low retrieval efficiency and long response time, failure to meet real-time interaction needs, difficulty in conducting in-depth trend analysis, collaborative analysis and comparative analysis, and lack of intelligent analysis report generation capabilities.

[0006] The first aspect of this application provides an intelligent intelligence analysis system based on multimodal technology topic trajectories, comprising: a homepage module for establishing indexes of at least one patent database and / or at least one paper database to retrieve patent data and / or paper data; a multimodal data search module for performing semantic retrieval of the target database based on the patent data and / or the paper data using a preset model context protocol to filter results that do not meet multi-dimensional conditions and obtain retrieval results; an intelligent analysis module for cleaning and standardizing the retrieval results to obtain final data that meets preset data quality and consistency conditions, and identifying the user's intelligence generation needs according to the analysis type selected by the user; an AI-driven engine for generating corresponding professional analysis reports based on the final data and the intelligence generation needs using a preset large language model; and a visualization display module for generating multi-dimensional data visualization information based on the professional analysis reports to display the professional analysis reports to the user.

[0007] Through the above-mentioned technical means, the embodiments of this application can establish a database index for retrieval, use a preset model context protocol for semantic retrieval and filter results, and after cleaning and standardization, combine user needs to generate professional analysis reports from a large language model and display them visually, thereby realizing efficient discovery, intelligent analysis and visualization of scientific and technological intelligence, effectively improving the efficiency and accuracy of intelligence analysis, enabling users to quickly obtain comprehensive and valuable scientific and technological intelligence, and providing strong support for decision-making.

[0008] Optionally, in one embodiment of this application, the homepage module includes: a function navigation submodule, used to generate at least one navigation area based on the at least one patent database and / or at least one paper database; a user management submodule, used to generate at least one daily dynamic area based on the at least one patent database and / or at least one paper database; a system configuration submodule, used to generate at least one technical service area based on the at least one patent database and / or at least one paper database; and a daily dynamic submodule, used to generate at least one research results area based on the at least one patent database and / or at least one paper database.

[0009] Through the above-mentioned technical means, the embodiments of this application can generate a navigation area, a daily dynamic area, a technical service area, and a research results area based on patent and / or paper databases. The multi-area index construction method makes the database structure clearer, facilitates users to quickly locate information according to different needs, and improves the convenience and relevance of data retrieval.

[0010] Optionally, in one embodiment of this application, the multimodal data search module includes: a patent search submodule, used for multi-field composite weight query, IPC classification, application year and relevance multi-dimensional filtering and sorting; a paper search submodule, used for multi-dimensional retrieval and filtering of title, abstract, journal and citation count; and an intelligent data acquisition engine based on model context protocol, used for semantic efficient retrieval of the ElasticSearch database based on model context protocol, supporting multi-dimensional condition filtering and high concurrency access.

[0011] Through the above-mentioned technical means, the embodiments of this application can perform multi-dimensional filtering and sorting, multi-dimensional retrieval and screening, and efficient retrieval of the ElasticSearch database based on the model context protocol. It supports multi-dimensional condition filtering and high-concurrency access, increases the accuracy and efficiency of retrieval, can accurately filter irrelevant information, and at the same time meet the stable operation in high-concurrency scenarios, ensuring that users can obtain the results they need in a timely manner.

[0012] Optionally, in one embodiment of this application, the intelligent analysis module includes: a patent analysis submodule for analyzing patent technology development trends, inventor and applicant distribution, and technological hotspots; a paper analysis submodule for analyzing research topics, author contributions, and citations; an author analysis submodule for analyzing the professional fields, innovation capabilities, and cooperation networks of research institutions; an industry-academia-research collaboration analysis submodule for evaluating the cooperative relationships between enterprises, universities, and research institutes; and an institution comparison analysis submodule for conducting multi-dimensional comparative analysis of different institutions or patent groups.

[0013] Through the above-mentioned technical means, the embodiments of this application can clean and standardize the search results, identify intelligence needs based on the analysis type selected by the user, and conduct multi-faceted analysis such as the development trend of patent technology, ensuring the high quality and consistency of the data, while accurately identifying the user's intelligence generation needs, laying a solid foundation for the subsequent generation of analysis reports that meet the user's expectations.

[0014] Optionally, in one embodiment of this application, the AI-driven engine includes: a prompting engineering submodule, used to automatically construct specialized prompts based on analysis needs; a data fusion submodule, used to integrate multi-source data to form an analysis basis; and a report generation submodule, used to automatically generate structured professional analysis reports, supporting various report types such as institutional theme trajectory, patents, papers, author academic profiles, industry-academia-research collaborations, regional analysis, and institutional comparisons.

[0015] Through the above-mentioned technical means, the embodiments of this application can utilize large language models to generate professional analysis reports, automatically construct prompt words, integrate multi-source data, and produce various types of structured reports, thereby achieving automated and professional report generation, covering multiple analysis dimensions, and meeting the intelligence needs of different scenarios.

[0016] Optionally, in one embodiment of this application, the visualization module includes: a trend chart submodule for displaying development trends over time; a relationship network submodule for visualizing the relationships and cooperation networks between entities; a distribution map submodule for displaying geographical distribution characteristics and innovation hotspots; and a comparison chart submodule for displaying comparative analysis results, including radar charts, heat maps, and timeline charts.

[0017] Through the above-mentioned technical means, the embodiments of this application can generate multi-dimensional visual information based on the report, displaying development trends, correlations and other content, making complex analysis results more intuitive and easy to understand, helping users to quickly grasp core information, and improving users' understanding and application of intelligence.

[0018] A second aspect of this application provides an intelligent intelligence analysis method based on multimodal technology topic trajectories, comprising: establishing an index for at least one patent database and / or at least one paper database to retrieve patent data and / or paper data; performing semantic retrieval on the target database based on the patent data and / or the paper data using a preset model context protocol to filter results that do not meet multi-dimensional conditions, thereby obtaining retrieval results; cleaning and standardizing the retrieval results to obtain final data that meets preset data quality and consistency conditions, and identifying the user's intelligence generation needs according to the analysis type selected by the user; generating a corresponding professional analysis report based on the final data and the intelligence generation needs using a preset large language model; and generating multi-dimensional data visualization information based on the professional analysis report to display the professional analysis report to the user.

[0019] Through the above-mentioned technical means, the embodiments of this application can establish a database index for retrieval, use a preset model context protocol for semantic retrieval and filter results, and after cleaning and standardization, combine user needs to generate professional analysis reports from a large language model and display them visually, thereby realizing efficient discovery, intelligent analysis and visualization of scientific and technological intelligence, effectively improving the efficiency and accuracy of intelligence analysis, enabling users to quickly obtain comprehensive and valuable scientific and technological intelligence, and providing strong support for decision-making.

[0020] Optionally, in one embodiment of this application, the establishment of an index to at least one patent database and / or at least one paper database for retrieving patent data and / or paper data includes: generating at least one navigation area based on the at least one patent database and / or at least one paper database; generating at least one daily dynamic area based on the at least one patent database and / or at least one paper database; generating at least one technical service area based on the at least one patent database and / or at least one paper database; and generating at least one research results area based on the at least one patent database and / or at least one paper database.

[0021] Through the above-mentioned technical means, the embodiments of this application can generate a navigation area, a daily dynamic area, a technical service area, and a research results area based on patent and / or paper databases. The multi-area index construction method makes the database structure clearer, facilitates users to quickly locate information according to different needs, and improves the convenience and relevance of data retrieval.

[0022] Optionally, in one embodiment of this application, the step of performing semantic retrieval on the target database based on the patent data and / or the paper data using a preset model context protocol to filter out results that do not meet the multi-dimensional conditions and obtain retrieval results includes: performing multi-field composite weighted queries, multi-dimensional filtering and sorting in IPC classification, application year, and relevance; performing multi-dimensional retrieval and filtering in title, abstract, journal, and citation count; and performing efficient semantic retrieval of the ElasticSearch database based on the model context protocol, supporting multi-dimensional condition filtering and high-concurrency access.

[0023] Through the above-mentioned technical means, the embodiments of this application can perform multi-dimensional filtering and sorting, multi-dimensional retrieval and screening, and efficient retrieval of the ElasticSearch database based on the model context protocol. It supports multi-dimensional condition filtering and high-concurrency access, increases the accuracy and efficiency of retrieval, can accurately filter irrelevant information, and at the same time meet the stable operation in high-concurrency scenarios, ensuring that users can obtain the results they need in a timely manner.

[0024] Optionally, in one embodiment of this application, the step of cleaning and standardizing the search results to obtain final data that meets preset data quality and consistency conditions, and identifying the user's intelligence generation needs according to the analysis type selected by the user, includes: conducting analysis on patent technology development trends, inventor and applicant distribution, and technology hotspots; conducting analysis on research topics, author contributions, and citations; conducting analysis on the professional fields, innovation capabilities, and cooperation networks of research institutions; conducting evaluation of cooperation relationships between enterprises, universities, and research institutes; and conducting multi-dimensional comparative analysis of different institutions or patent groups.

[0025] Through the above-mentioned technical means, the embodiments of this application can clean and standardize the search results, identify intelligence needs based on the analysis type selected by the user, and conduct multi-faceted analysis such as the development trend of patent technology, ensuring the high quality and consistency of the data, while accurately identifying the user's intelligence generation needs, laying a solid foundation for the subsequent generation of analysis reports that meet the user's expectations.

[0026] Optionally, in one embodiment of this application, the step of generating a corresponding professional analysis report based on the final data and the intelligence generation requirements using a preset large language model includes: automatically constructing specialized prompt words according to the analysis requirements; integrating multi-source data to form an analysis basis; and automatically generating a structured professional analysis report that supports various report types such as institutional theme trajectory, patents, papers, author academic profiles, industry-university-research cooperation, regional analysis, and institutional comparison.

[0027] Through the above-mentioned technical means, the embodiments of this application can utilize large language models to generate professional analysis reports, automatically construct prompt words, integrate multi-source data, and produce various types of structured reports, thereby achieving automated and professional report generation, covering multiple analysis dimensions, and meeting the intelligence needs of different scenarios.

[0028] Optionally, in one embodiment of this application, the step of generating multi-dimensional data visualization information based on the professional analysis report to display the professional analysis report to the user includes: displaying development trends over time; visualizing the relationships and cooperation networks between entities; displaying geographical distribution characteristics and innovation hotspots; and displaying comparative analysis results, including radar charts, heat maps, and timeline charts.

[0029] Through the above-mentioned technical means, the embodiments of this application can generate multi-dimensional visual information based on the report, displaying development trends, correlations and other content, making complex analysis results more intuitive and easy to understand, helping users to quickly grasp core information, and improving users' understanding and application of intelligence.

[0030] A third aspect of this application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the intelligent intelligence analysis method based on multimodal technology theme trajectories as described in the above embodiments.

[0031] A fourth aspect of this application provides a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described intelligent intelligence analysis method based on multimodal technology topic trajectories.

[0032] A fifth aspect of this application provides a computer program product that stores a computer program that, when executed by a processor, implements the above-described intelligent intelligence analysis method based on multimodal technology topic trajectories.

[0033] This application embodiment can establish a database index for retrieval, utilize a preset model context protocol for semantic retrieval and filter results, and after cleaning and standardization, combine user needs to generate professional analysis reports from a large language model and visualize them. This achieves efficient discovery, intelligent analysis, and visualization of scientific and technological intelligence, effectively improving the efficiency and accuracy of intelligence analysis, allowing users to quickly obtain comprehensive and valuable scientific and technological intelligence, and providing strong support for decision-making. Therefore, it solves the problems in related technologies, such as supporting only a single data source, resulting in the inability to integrate multimodal data for comprehensive analysis, low retrieval efficiency and long response time, inability to meet real-time interaction needs, difficulty in conducting in-depth trend analysis, collaborative analysis, and comparative analysis, and lack of intelligent analysis report generation capabilities.

[0034] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0035] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0036] Figure 1 This is a schematic diagram of the structure of an intelligent intelligence analysis system based on multimodal technology topic trajectories provided in an embodiment of this application;

[0037] Figure 2 This is a schematic diagram of an intelligent intelligence analysis system architecture based on multimodal technology data according to an embodiment of this application;

[0038] Figure 3 This is a flowchart illustrating the workflow of a multimodal data search module according to an embodiment of this application.

[0039] Figure 4 A flowchart illustrating the report generation process of an AI-driven engine according to an embodiment of this application;

[0040] Figure 5 This is a flowchart of industry-academia-research collaboration analysis provided according to one embodiment of this application;

[0041] Figure 6 A flowchart of mechanism technical subject trajectory analysis provided according to one embodiment of this application;

[0042] Figure 7A flowchart for generating user interaction and analysis according to an embodiment of this application;

[0043] Figure 8 This is a flowchart illustrating an intelligent intelligence analysis method based on multimodal technology topic trajectories provided in an embodiment of this application.

[0044] Figure 9 This is a schematic diagram of the structure of an electronic device provided according to an embodiment of this application.

[0045] Figure label:

[0046] 10-Intelligent intelligence analysis method and device based on multimodal technology theme trajectory; 100-Home module, 200-Multimodal data search module, 300-Intelligent analysis module, 400-AI driving engine, 500-Visualization display module; 901-Memory, 902-Processor, 903-Communication interface. Detailed Implementation

[0047] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0048] The following description, with reference to the accompanying drawings, describes an intelligent intelligence analysis system and method based on multimodal science and technology theme trajectories, according to embodiments of this application. Addressing the problems mentioned in the background art, which only support a single data source, resulting in the inability to integrate multimodal data for comprehensive analysis, low retrieval efficiency, long response times, inability to meet real-time interaction needs, difficulty in conducting in-depth trend analysis, collaborative analysis, and comparative analysis, and lack of intelligent analysis report generation capabilities, this application provides an intelligent intelligence analysis system based on multimodal science and technology theme trajectories. In this system, a database index can be established for retrieval, semantic retrieval using a preset model context protocol is employed, and results are filtered. After cleaning and standardization, combined with user needs, a large language model generates professional analysis reports and displays them visually. This achieves efficient discovery, intelligent analysis, and visualization of science and technology intelligence, effectively improving the efficiency and accuracy of intelligence analysis, allowing users to quickly obtain comprehensive and valuable science and technology intelligence, and providing strong support for decision-making. Therefore, this solves the problems in related technologies, such as the inability to integrate multimodal data for comprehensive analysis, low retrieval efficiency, long response times, inability to meet real-time interaction needs, difficulty in conducting in-depth trend analysis, collaborative analysis, and comparative analysis, and lack of intelligent analysis report generation capabilities.

[0049] Specifically, Figure 1This is a schematic diagram of the structure of an intelligent intelligence analysis system based on multimodal technology topic trajectories provided in an embodiment of this application.

[0050] like Figure 1 As shown, the intelligent intelligence analysis system 10 based on multimodal technology theme trajectories includes:

[0051] Homepage module 100 is used to create indexes for at least one patent database and / or at least one paper database to retrieve patent data and / or paper data.

[0052] It is understood that the index in this application embodiment can be used to structurally mark and store data (such as titles, abstracts, classification numbers, publication dates, etc.) in patent databases and paper databases.

[0053] In actual implementation, the embodiments of this application can establish efficient indexes for patent databases, paper databases, and patent databases, optimize data organization structure, and achieve this through the front-end Svelte framework. It supports responsive design, can adapt to different screen sizes, and ensures a good user experience.

[0054] The embodiments of this application can establish an index to enable rapid retrieval of patent and paper data, significantly shortening the data acquisition time, while providing a structured data source for subsequent multi-dimensional analysis of the system, thereby improving the efficiency of overall intelligence analysis.

[0055] Optionally, in one embodiment of this application, the homepage module 100 includes: a function navigation submodule for generating at least one navigation area based on at least one patent database and / or at least one paper database; a user management submodule for generating at least one daily dynamic area based on at least one patent database and / or at least one paper database; a system configuration submodule for generating at least one technical service area based on at least one patent database and / or at least one paper database; and a daily dynamic submodule for generating at least one research results area based on at least one patent database and / or at least one paper database.

[0056] For example, in this embodiment, the homepage module can be implemented using a home component, including a function navigation area, a daily news area, a technical service area, and a research results area. The function navigation area provides entry buttons for patent search, paper search, industry-academia collaboration, and patent and paper institution comparison; users can click to switch to the corresponding function module. The daily news area provides the latest paper updates, including titles and abstracts; clicking "Read More" allows users to read the full text of the paper. The technical service area provides information on the relevant technical applications involved in this system, as well as the services offered. The research results area provides the latest patent inventions, etc. The system adopts a responsive design, adapting to different screen sizes to ensure a good user experience.

[0057] The embodiments of this application can make the homepage functions clearer by refining the division of labor, allowing users to quickly find the functions they need and obtain the latest data. At the same time, the system operating parameters can be flexibly configured, improving the system's ease of use and adaptability.

[0058] The multimodal data search module 200 is used to perform semantic retrieval of the target database based on patent data and / or paper data, using a preset model context protocol to filter out results that do not meet the multi-dimensional conditions and obtain the search results.

[0059] It is understood that in the embodiments of this application, the preset model context protocol can be a rule for defining the processing of retrieval requests by a large language model, such as MCP (Model Context Protocol). Semantic retrieval can be an upgrade relative to traditional keyword retrieval, which can understand the deeper meaning of the text.

[0060] In actual implementation, the embodiments of this application can encapsulate complex query logic, with SearchBox providing the search input interface, Facets implementing multi-dimensional filtering, and SearchResults responsible for displaying the results. At the same time, the system uses MCP technology to optimize the connection efficiency with ElasticSearch, enabling efficient access to the patent database and paper database.

[0061] Specifically, this embodiment can first initialize the configuration, load the corresponding retrieval configuration according to the search type (patent or paper), including API (Application Programming Interface) endpoints, index names, and weighting strategies; then construct a compound query statement, supporting multi-field, multi-condition filtering and highlighting, accessing the patent database and paper database, performing multi-dimensional search sorting, executing the retrieval, and employing different caching strategies for different types of queries. For high-frequency queries, a memory caching mechanism is used to achieve millisecond-level response; for complex queries, result prefetching and asynchronous loading are used to ensure a good interactive experience; using the search result parsing method, MCP technology is used to efficiently interact with the ElasticSearch database, responsible for parsing the data returned by ElasticSearch, processing ordinary fields and highlighted fields, and constructing structured result objects; finally, the results are displayed by rendering the search results, showing patent or paper information in card form, and providing pagination control.

[0062] The embodiments of this application can overcome the limitations of keyword matching through semantic search, and combine multi-dimensional filtering to exclude irrelevant data, significantly improving the accuracy of search results and allowing users to quickly obtain information that is highly relevant to their needs.

[0063] Optionally, in one embodiment of this application, the multimodal data search module 200 includes: a patent search submodule for performing multi-field composite weighted queries, IPC (International Patent Classification) classification, multi-dimensional filtering and sorting based on application year and relevance; a paper search submodule for performing multi-dimensional retrieval and filtering based on title, abstract, journal, and citation count; and an intelligent data acquisition engine based on the model context protocol for semantically efficient retrieval of the ElasticSearch database based on the model context protocol, supporting multi-dimensional conditional filtering and high-concurrency access.

[0064] In actual implementation, the patent search submodule in this embodiment can be used for multi-field composite weighted queries, IPC classification, application year, and multi-dimensional filtering and sorting based on relevance. Specifically, this embodiment can construct complex ElasticSearch query statements by creating an ElasticSearch search method, supporting full-text patent search and querying multiple fields based on searchTerm. It also supports multi-dimensional filtering, such as filtering based on IPC classification, legal status, and time range, as well as sorting control, such as relevance sorting and time sorting, and highlighting, adding highlight marks to matched content.

[0065] Furthermore, the paper search submodule allows for multi-dimensional searching and filtering based on title, abstract, journal, and citation count. It strengthens the weighting of academic indicators such as journal ranking and citation count, employing a multi-field composite weighting query strategy. For example, in paper retrieval, the title weight is set to 3, the abstract weight to 2, and the publishing journal weight to 1.5. The paper search submodule supports multi-dimensional filtering, such as filtering by time range, document type, and citation count. It supports sorting controls, such as relevance sorting, time sorting, and citation count sorting, and also supports highlighting, adding highlight marks to matched content.

[0066] This intelligent data acquisition engine, based on the Model Context Protocol (MTP), enables semantic and efficient retrieval of Elasticsearch databases, supporting multi-dimensional filtering and high-concurrency access. It sends POST requests to the Elasticsearch server via API, with the request body containing the constructed query statement. The Facets component provides rich filtering capabilities, allowing users to filter by year, document type, citation count, and other dimensions. When filtering conditions change, the system reconstructs the query and executes the retrieval, updating the results list in real time. This multi-dimensional query capability enables users to quickly locate subsets of data of interest, providing a foundation for subsequent in-depth analysis.

[0067] This application embodiment can perform multi-dimensional filtering and sorting through multi-field composite weight queries, multi-dimensional retrieval and screening, and efficient retrieval of the ElasticSearch database based on the model context protocol. It supports multi-dimensional condition filtering and high-concurrency access, increases the accuracy and efficiency of retrieval, can accurately filter irrelevant information, and at the same time meet the stable operation in high-concurrency scenarios, ensuring that users can obtain the results they need in a timely manner.

[0068] The intelligent analysis module 300 is used to clean and standardize the search results to obtain the final data that meets the preset data quality and consistency conditions, and to identify the user's intelligence generation needs based on the analysis type selected by the user.

[0069] It is understood that in the embodiments of this application, cleaning and standardization can be used to remove duplicate and erroneous values ​​from the data and to unify the data format; the preset data quality and consistency conditions can be data integrity (such as patent data needing to include core fields such as application date, applicant, and abstract, and paper data needing to include necessary information such as title, author, and journal) and data accuracy (such as patent numbers conforming to international or national unified coding rules, and paper citation formats conforming to standards such as GB / T7714). The preset data quality and consistency conditions can be set by those skilled in the art according to the actual situation, and no specific restrictions are imposed here.

[0070] For example, embodiments of this application can achieve intelligent analysis through FloatingPanel and Chat components. The FloatingPanel component provides methods such as report generation, institution analysis, and author analysis to trigger different types of analysis tasks. When a user selects an analysis object, the system will construct specialized prompts, such as "Please analyze the institution {institutionName}, AffiliationID={institutionId}", and then call the Chat component's automated AI report generation method to generate an analysis report.

[0071] The embodiments of this application can ensure data quality through data cleaning and standardization, and accurately match analysis directions through demand identification, laying the foundation for generating reliable intelligence reports that meet user needs in the future.

[0072] Optionally, in one embodiment of this application, the intelligent analysis module 300 includes: a patent analysis submodule for analyzing patent technology development trends, inventor and applicant distribution, and technological hotspots; a paper analysis submodule for analyzing research topics, author contributions, and citations; an author analysis submodule for analyzing the professional fields, innovation capabilities, and cooperation networks of research institutions; an industry-academia-research collaboration analysis submodule for evaluating the cooperative relationships between enterprises, universities, and research institutes; and an institution comparison analysis submodule for conducting multi-dimensional comparative analysis of different institutions or patent groups.

[0073] In actual implementation, the embodiments of this application can use the patent analysis submodule to statistically analyze the number of patent applications in each year and draw trend charts, identify core R&D personnel by the frequency of inventors, judge the industry competitive landscape by the distribution of applicants, and conduct analysis on patent technology development trends, inventor and applicant distribution, and technology hotspots; the paper analysis submodule can identify research topics by keyword clustering (such as the sub-topics of "carbon capture" and "new energy" in papers related to "carbon neutrality"), and statistically analyze the number of articles published and the frequency of citations by authors to evaluate their contributions for analysis of research topics, author contributions, and citations; the author analysis submodule is used to analyze the professional fields, innovation capabilities, and cooperation networks of research institutions.

[0074] The industry-academia-research collaboration analysis submodule can acquire information such as the research direction or abstract of the paper that the user inputs on the front end, indicating the desired collaboration. It also acquires data on the patent applicant type, region, IPC classification number, and application year range specified by the user, as well as any special collaboration analysis requirements. Based on patent-related data in the ElasticSearch database, it performs queries according to the user-selected patent applicant type, region, IPC classification number, and application year. Using the MCP method, the queried data is transmitted to a local AI large-scale language model, and combined with relevant specialized prompts, a large-scale language model is used to conduct in-depth data analysis, calculate key network indicators, and evaluate the collaborative relationships between enterprises, universities, and research institutions. This includes: industry-academia-research collaboration patent output trend analysis, industry-academia-research collaboration technology field distribution analysis, industry-academia-research collaboration network analysis, industry-academia-research collaboration patent quality assessment, and industry-academia-research collaboration technology transfer efficiency analysis.

[0075] The institutional comparison analysis submodule is used for multi-dimensional comparative analysis of different institutions or patent groups. Users input the institution name, and the system uses the Elasticsearch database to search for relevant patent and paper institution names, returning these results as input suggestions to the front end. Users can then select two target institutions from these suggestions for comparative analysis. For the selected two institutions, the system efficiently retrieves their patent and paper data from the Elasticsearch database in parallel. This data covers multiple key dimensions, including but not limited to technology distribution, time trends, collaboration networks, and citation information.

[0076] Specifically, this application embodiment inputs the acquired multimodal data and preset specialized prompts (e.g., "Please compare and analyze the similarities and differences between the following two institutions in terms of innovation output, technology distribution, citation influence, and cooperation network, and provide key findings and development suggestions") into a local AI large language model. The AI ​​model will automatically perform a series of complex indicator calculations and in-depth analyses, including comparing innovation output, quantifying and evaluating the differences between the two institutions in the quantity and quality of patents and papers (such as high-value patents and highly cited papers). It will also reveal the similarities and differences and focuses of the two institutions in core technology fields and research directions by analyzing IPC classifications and research keywords. Through the time trends and development trajectories of the two institutions, it analyzes the time series data of patent applications and paper publications to gain insights into the development trends, breakthroughs, and transformations of the two institutions' innovation activities. Through their citation influence and academic / technical status, it compares patent citations and paper citations to assess the academic influence and technological leadership of the two institutions in their respective fields. Through their cooperation networks and collaborative efficiency, it analyzes the partner groups, cooperation intensity, and cooperation models of the two institutions to reveal their collaborative innovation capabilities and analyze their competitive landscape and advantages and disadvantages. AI will integrate the above indicators to intelligently identify the relative strengths and weaknesses of the two organizations, as well as potential competitive threats or cooperation opportunities.

[0077] Furthermore, a structured, professional comparative analysis report is automatically generated. The patent report will include: comparative analysis of the institution's patent output trends, comparative analysis of the institution's technological field distribution, comparative analysis of the institution's innovation teams, comparative assessment of the institution's patent quality, and comparative analysis of the institution's technology layout. The paper report will include: basic information and geographical location of the institution, the institution's annual paper output and citation trends, the institution's research field distribution, cooperation network, innovation index, and influence indicators. The institution comparative analysis function supports two modes: patent institution comparison and paper institution comparison. These two modes complement each other: patent institution comparison focuses on technological innovation capabilities, patent portfolio strategies, and assessment of technology commercialization value; paper institution comparison focuses on academic influence, research theme evolution, and talent team development. Together, these two modes provide a comprehensive profile of the institution's innovation capabilities.

[0078] This application embodiment can clean and standardize the search results, identify intelligence needs based on the analysis type selected by the user, and conduct multi-faceted analysis such as patent technology development trends to ensure high-quality and consistent data. At the same time, it can accurately identify the user's intelligence generation needs, laying a solid foundation for generating analysis reports that meet the user's expectations.

[0079] AI-driven engine 400 is used to generate corresponding professional analysis reports based on the final data and intelligence generation requirements using a pre-set large language model.

[0080] It is understood that the preset large language model in the embodiments of this application can be an artificial intelligence model with text understanding and generation capabilities, such as GPT-4, LLaMA, etc., which can generate analysis reports that conform to professional logic based on input data and requirements. The preset large language model can be set by those skilled in the art according to the actual situation, and no specific restrictions are made here.

[0081] In practical implementation, this embodiment can be achieved through the Chat component, comprising three main steps: prompt word construction, data fusion, and report generation. The system dynamically constructs optimized prompt words based on the type of analysis object (institution, patent, paper, etc.) and the analysis objective, and acquires relevant data in real time using MCP technology. Post-processing of the generated report includes structural verification, data consistency checks, and format optimization to ensure the report's professionalism and readability. The report types supported by this embodiment include: institution theme trajectory report, patent analysis report, paper analysis report, author academic profile analysis, industry-academia-research collaboration analysis, regional analysis report, and institution comparative analysis report. Different types of reports employ differentiated prompt templates and data fusion strategies to ensure the targeted and professional nature of the generated content. Tests show that the average generation time for institution theme trajectory reports is approximately 30 seconds, author academic profile analysis is approximately 60 seconds, patent applicant analysis is 30-60 seconds, and patent inventor analysis is approximately 50 seconds.

[0082] This application embodiment can deeply integrate large-scale language models with professional data to automatically generate high-quality professional analysis reports, thereby achieving automated generation of professional analysis reports and improving intelligence production efficiency.

[0083] Optionally, in one embodiment of this application, the AI-driven engine 400 includes: a prompting engineering submodule for automatically constructing specialized prompts based on analysis needs; a data fusion submodule for integrating multi-source data to form an analysis basis; and a report generation submodule for automatically generating structured professional analysis reports, supporting various report types such as institutional theme trajectories, patents, papers, author academic profiles, industry-academia-research collaborations, regional analysis, and institutional comparisons.

[0084] For example, in the embodiments of this application, the prompting engineering submodule can automatically construct specialized prompts based on the analysis object and target. For example, for paper analysis, the prompt format is "Please generate an analysis report for the paper "{paperTitle}", PaperID = {paperId}\nAbstract: {abstract}"; for institution analysis, the prompt format is "Please analyze the institution\"{institutionName}\", AffiliationID = {institutionId||'Unknown ID'}".

[0085] The data fusion submodule can call the method for generating automated analysis reports. First, it automatically parses the prompt words, extracts key entities and attributes, and retrieves relevant supplementary data from Elasticsearch based on the entity information. Then, it merges the search results with the original prompt words to build an enhanced context and integrate multi-source data to form the basis for analysis.

[0086] The report generation submodule supports various report types, including institutional thematic trajectories, patents, papers, author academic profiles, industry-academia-research collaborations, regional analysis, and institutional comparisons. This embodiment uses enhanced context as input and invokes a large language model to generate a multi-part structured report. First, it performs an abstract to summarize the core characteristics and main findings of the analyzed object. Then, it conducts data analysis, performing quantitative analysis based on retrieved data. Next, it provides trend insights to identify development trends and turning points, assesses competitive landscape, and evaluates relative advantages and challenges. Finally, it provides development recommendations, offering data-driven strategic suggestions. The system automatically generates a structured, professional analysis report, including an abstract, data analysis, trend insights, competitive landscape, and development recommendations, and performs structural validation, data consistency checks, and format optimization on the report.

[0087] This application embodiment can utilize a large language model to generate professional analysis reports, automatically construct prompt words, integrate multi-source data, and produce various types of structured reports, thereby achieving automated and professional report generation, covering multiple analysis dimensions, and meeting the intelligence needs of different scenarios.

[0088] The visualization module 500 is used to generate multi-dimensional data visualization information based on professional analysis reports to present the professional analysis reports to users.

[0089] It is understood that the multi-dimensional data visualization information in the embodiments of this application can be the transformation of the numbers and text descriptions in professional analysis reports into intuitive charts and graphs, such as line graphs and network diagrams.

[0090] For example, embodiments of this application can automatically generate trend charts, relationship networks, distribution maps, and comparison charts based on AI analysis results; support multi-dimensional interaction and dynamic filtering to improve users' efficiency in understanding complex data; and the visualization results support highlighting key nodes, areas of strength, and potential cooperation opportunities. Furthermore, the AI ​​will intelligently recommend the chart type that best reflects the comparison results and may highlight key differences within the chart.

[0091] The embodiments of this application can transform complex analysis results into intuitive and easy-to-understand graphics through data visualization, reducing the time cost for users to interpret data and improving the efficiency of understanding the core conclusions of intelligence.

[0092] Optionally, in one embodiment of this application, the visualization module 500 includes: a trend chart submodule for displaying development trends over time; a relationship network submodule for visualizing the relationships and cooperation networks between entities; a distribution map submodule for displaying distribution characteristics and innovation hotspots in geographic space; and a comparison chart submodule for displaying comparative analysis results, including radar charts, heat maps, and timeline charts.

[0093] It is understood that, in the embodiments of this application, a radar chart can be a graph that connects the data values ​​of different subjects in each dimension in the form of polygons through multiple coordinate axes (each axis represents an analysis dimension) starting from the same point; a heat map can be a graph that reflects the density and magnitude of data in a specific area or matrix through the shade of color; and a timeline chart can be a graph that arranges data in chronological order, with the horizontal axis representing time and the vertical axis representing data values, and displays the changes of data over time in the form of lines, bars, etc.

[0094] In actual implementation, the trend chart submodule in this application embodiment can display the development trend over time, such as using a stacked bar chart to show "the annual distribution of patents in a certain field by technology branch", clearly presenting the time evolution of each branch.

[0095] The relationship network submodule presents the cooperation network through a force-directed graph. In the graph, node size intuitively represents the influence of an organization, edge width represents the strength of cooperation, and node color is used to distinguish the type of organization (enterprise, university, research institute), thus visualizing the relationships and cooperation networks between entities. Furthermore, based on AI analysis results, the system will automatically highlight key nodes, advantageous cooperation areas, and potential cooperation opportunities, improving users' efficiency in understanding complex networks.

[0096] The distribution map submodule can display the distribution characteristics and innovation hotspots in geospatial space, link regional data with specific geographical areas, and intuitively present the spatial distribution patterns and innovation resource aggregation characteristics involved in scientific and technological intelligence, allowing users to quickly grasp the differences and characteristics of different regions in terms of technology research and development, output results, etc.

[0097] The comparative charts submodule is used to display the results of comparative analysis, including radar charts, heatmaps, and timeline charts. Radar charts clearly show the comparison of the two institutions' capabilities across multiple dimensions (such as innovation output, citation impact, and technological breadth). Heatmaps visually reflect the distribution density and differences of the two institutions across different technological fields or research topics. Timeline charts dynamically present the development trajectory and key milestones of the two institutions' innovation output.

[0098] The embodiments of this application can generate multi-dimensional visual information based on reports, displaying development trends, relationships, and other content, making complex analysis results more intuitive and easy to understand, helping users quickly grasp core information, and improving users' understanding and application of intelligence.

[0099] Specifically, it can be combined with Figures 2 to 7 As shown, the working principle of the intelligent intelligence analysis system based on multimodal technology theme trajectory in this application is explained in detail with a specific embodiment.

[0100] Figure 2 This is a schematic diagram of an intelligent intelligence analysis system architecture based on multimodal technology data provided according to an embodiment of this application.

[0101] like Figure 2 As shown, this embodiment of the application can adopt a front-end microservice architecture and a back-end distributed computing framework. The core modules include a homepage module, a search module, an analysis module, and an AI report engine. The system front-end is implemented based on the Svelte framework and mainly includes three core components: home, FloatingPanel, and Chat. The home component provides system function navigation, the FloatingPanel component implements multimodal search and result display, and the Chat component is responsible for AI-driven report generation.

[0102] Figure 3 This is a flowchart illustrating the workflow of a multimodal data search module provided according to an embodiment of this application.

[0103] like Figure 3 As shown, embodiments of this application may include the following steps:

[0104] Step S301: Search module: Patent data and paper data.

[0105] In this embodiment of the application, relevant patent data and paper data can be searched using the search module.

[0106] Step S302: Initialize the search.

[0107] In this embodiment, the retrieval configuration can be initialized, and API endpoints, index names, and weight strategies can be loaded.

[0108] Step S303: Compound query.

[0109] In this application embodiment, a composite query statement can be constructed, supporting multi-field, multi-condition filtering and highlighting.

[0110] Step S304: MCP connects to the database.

[0111] In this embodiment, the MCP intelligent data acquisition engine can efficiently interact with the ElasticSearch database.

[0112] Step S305: Obtain the corresponding data.

[0113] In this application embodiment, corresponding data and search results can be obtained.

[0114] Figure 4 A flowchart illustrating the report generation process of an AI-driven engine provided according to an embodiment of this application.

[0115] like Figure 4 As shown, embodiments of this application may include the following steps.

[0116] Step S401: Select the object for which trajectory analysis is required.

[0117] In this embodiment, the user can identify the object and type selected for trajectory analysis and trigger the analysis task.

[0118] Step S402: Use MCP to obtain database data.

[0119] In this embodiment, MCP can be used to obtain database data, automatically construct specialized prompts, and extract key entities and attributes.

[0120] Step S403: Analyze the data based on the prompt words.

[0121] In this embodiment of the application, the data can be analyzed based on prompt words, relevant supplementary data can be retrieved and fused to construct an enhanced context.

[0122] Step S404: Generate AI large model report.

[0123] In this application embodiment, a large language model can be invoked to generate a structured analysis report, including a summary, data analysis, trend insights, competitive landscape, and development recommendations.

[0124] Step S405: Inspection report, generate complete page.

[0125] In this application embodiment, the report can be structurally verified, data consistency checked, and format optimized.

[0126] Figure 5 This is a flowchart of industry-academia-research collaborative analysis provided according to one embodiment of this application.

[0127] like Figure 5 As shown, this application embodiment can generate analysis reports among enterprises, universities, and research institutions to assess the health of the regional innovation ecosystem and provide decision support for precise industrial upgrading. The analysis process includes the following steps:

[0128] Step S501: Data preparation: including paper abstract, research direction, patent type, region, year, and special requirements.

[0129] In this embodiment, the system can acquire information such as the research direction or abstract of the paper that the user inputs on the front end, indicating the desired collaboration. It can also acquire data on the patent applicant type, region, IPC classification number, and application year range specified by the user, as well as any specific collaboration analysis requirements.

[0130] Step S502: Use MCP to obtain database data.

[0131] In this embodiment, a collaborative network can be constructed. Based on patent-related data in the ElasticSearch database, queries can be performed according to user-selected conditions such as patent applicant type, region, IPC classification number, and application year. The queried data is then transmitted to the local AI large language model via MCP to prepare for subsequent analysis.

[0132] Step S503: Analyze the data based on the prompt words.

[0133] In this embodiment, the retrieved data can be used as input to an AI large language model, combined with preset specialized prompts. The large language model will perform in-depth analysis of the data and calculate key network indicators, including: industry-academia collaboration patent output trend analysis, industry-academia collaboration technology field distribution analysis, industry-academia collaboration network analysis, industry-academia collaboration patent quality assessment, and industry-academia collaboration technology transfer efficiency analysis.

[0134] Step S504: Generate a report from the large AI model.

[0135] In this embodiment, all network analysis results generated by the AI ​​large language model can be used as input, and the AI-driven report generation function can be invoked to automatically generate a collaborative analysis report between industry, academia and research.

[0136] Step S505: Inspection report, generate complete page.

[0137] In this application embodiment, the report can be structurally verified, data consistency checked, and format optimized.

[0138] This application embodiment provides users with a macro perspective through the industry-university-research collaborative analysis function, helping government departments understand the regional innovation ecosystem, assisting enterprises in finding potential partners, and supporting universities in optimizing their industry-university-research cooperation strategies.

[0139] Figure 6 This is a flowchart of a mechanism technical subject trajectory analysis according to an embodiment of this application.

[0140] like Figure 6 As shown, this application embodiment can support multi-dimensional, AI-driven intelligent comparative analysis of patent institutions and paper institutions, helping users to deeply understand the differences between different institutions in terms of technical routes, innovation capabilities, and research focuses, providing precise support for competitive analysis and strategic cooperation decisions. The comparative analysis process is as follows:

[0141] Step S601: Enter the organization name.

[0142] In this embodiment, the organization name entered by the user can be identified.

[0143] Step S602: Elasticsearch returns an input prompt.

[0144] In this embodiment, the system can search for relevant patent and paper institution names in the Elasticsearch database and return these results as input suggestions to the front end. Users can then select two target institutions from these intelligent suggestions for comparative analysis.

[0145] Step S603: Use MCP to obtain database data.

[0146] In this embodiment, patent and publication data can be retrieved from Elasticsearch databases in parallel and efficiently for two selected institutions. This data covers multiple key dimensions, including but not limited to technology distribution, time trends, collaboration networks, and citation information.

[0147] Step S604: Analyze the data based on the prompt words.

[0148] In this embodiment, the acquired multimodal data and preset specialized prompts (e.g., "Please compare and analyze the similarities and differences between the following two institutions in terms of innovation output, technology distribution, citation influence, and collaborative networks, and provide key findings and development suggestions") can be input into a local AI large language model. The AI ​​model will automatically perform a series of complex indicator calculations and in-depth analyses, including comparing innovation output, quantifying and evaluating the differences between the two institutions in the quantity and quality of patents and papers (e.g., high-value patents, highly cited papers). By analyzing IPC classifications and research keywords, it reveals the similarities and differences and focuses of the two institutions in core technology areas and research directions. By analyzing the time trends and development trajectories of the two institutions, it analyzes the time series data of patent applications and paper publications to gain insights into the development trends, breakthroughs, and transformations of the two institutions' innovation activities. By comparing patent citations and paper citations based on their citation influence and academic / technical status, it assesses the academic influence and technological leadership of the two institutions in their respective fields. By analyzing their collaborative networks and synergy efficiency, it analyzes the partner groups, collaboration intensity, and collaboration models of the two institutions to reveal their collaborative innovation capabilities and analyze their competitive landscape and strengths and weaknesses. AI will integrate the above indicators to intelligently identify the relative strengths and weaknesses of the two organizations, as well as potential competitive threats or cooperation opportunities.

[0149] Furthermore, embodiments of this application can intelligently generate and optimize various visualization charts based on the analysis results of the AI ​​large language model, intuitively presenting the comparison results. Radar charts clearly demonstrate the comparison of the two institutions' capabilities across multiple dimensions (such as innovation output, citation impact, and technological breadth). Heatmaps intuitively reflect the distribution density and differences of the two institutions across different technological fields or research topics. Timeline charts dynamically present the development trajectory of the two institutions' innovation output and a comparison of key milestones. In addition, AI will intelligently recommend the chart type that best reflects the comparison results and may highlight key differences in the charts.

[0150] Step S605: Generate a report from the large AI model.

[0151] In this embodiment, all comparative analysis results generated by the AI ​​large language model can be used as input, and an AI-driven report generation function can be invoked to automatically generate a structured professional comparative analysis report. The patent report will include: comparative analysis of the institution's patent output trends, comparative analysis of the institution's technological field distribution, comparative analysis of the institution's innovation teams, comparative evaluation of the institution's patent quality, and comparative analysis of the institution's technological layout. The paper report will include: basic information and geographical location of the institution, annual paper output and citation trends of the institution, distribution of the institution's research fields, cooperation network, innovation index, and influence indicators, etc.

[0152] Specifically, the institutional comparison analysis function supports two modes: patent institution comparison and paper institution comparison. These two modes complement each other: patent institution comparison focuses on technological innovation capabilities, patent portfolio strategies, and assessment of technology commercialization value; while paper institution comparison focuses on academic influence, evolution of research topics, and talent development. Together, these two modes provide a comprehensive profile of an institution's innovation capabilities.

[0153] Step S606: Inspection report, generate complete page.

[0154] In this application embodiment, the report can be structurally verified, data consistency checked, and format optimized.

[0155] Figure 7 A flowchart for generating user interaction and analysis according to an embodiment of this application is provided.

[0156] like Figure 7 As shown, embodiments of this application may include the following steps:

[0157] Step S701: AI large model analysis.

[0158] In this application embodiment, a large-scale language model can be used to generate a professional analysis report based on optimized prompt words and fused data.

[0159] Step S702: Generate a visualization chart.

[0160] In this application embodiment, trend charts, relationship networks, distribution maps, and comparison charts can be automatically generated based on AI analysis results.

[0161] Step S703: Full-range display.

[0162] In this application, the analysis results can be displayed intuitively using charts, network diagrams, and maps.

[0163] Step S704: Highlight key data and filter dynamically.

[0164] The embodiments of this application can support multi-dimensional interaction and dynamic filtering, improve the efficiency of users' understanding of complex data, and support the highlighting of key nodes, advantageous areas and potential cooperation opportunities.

[0165] The intelligent intelligence analysis system based on multimodal science and technology topic trajectories proposed in this application can establish a database index for retrieval, utilize a preset model context protocol for semantic retrieval and filter results, and after cleaning and standardization, combine user needs to generate professional analysis reports from a large language model and visualize them. This achieves efficient discovery, intelligent analysis, and visualization of science and technology intelligence, effectively improving the efficiency and accuracy of intelligence analysis, allowing users to quickly obtain comprehensive and valuable science and technology intelligence, and providing strong support for decision-making. Therefore, it solves the problems in related technologies that only support a single data source, resulting in the inability to integrate multimodal data for comprehensive analysis, low retrieval efficiency, long response time, inability to meet real-time interaction needs, difficulty in conducting in-depth trend analysis, collaborative analysis, and comparative analysis, and lack of intelligent analysis report generation capabilities.

[0166] Next, referring to the accompanying drawings, a method for intelligent intelligence analysis based on multimodal technology topic trajectories proposed according to embodiments of this application is described.

[0167] Figure 8 This is a flowchart illustrating an intelligent intelligence analysis method based on multimodal technology topic trajectories provided in an embodiment of this application.

[0168] like Figure 8 As shown, this intelligent intelligence analysis method based on multimodal technology topic trajectories includes the following steps:

[0169] In step S801, an index is created for at least one patent database and / or at least one paper database to retrieve patent data and / or paper data.

[0170] In step S802, based on patent data and / or paper data, a semantic search is performed on the target database using a preset model context protocol to filter out results that do not meet the multi-dimensional conditions, thereby obtaining the search results.

[0171] In step S803, the search results are cleaned and standardized to obtain final data that meets the preset data quality and consistency conditions, and the user's intelligence generation needs are identified according to the analysis type selected by the user.

[0172] In step S804, a corresponding professional analysis report is generated based on the final data and intelligence generation requirements using a preset large language model.

[0173] In step S805, multi-dimensional data visualization information is generated based on the professional analysis report to present the professional analysis report to the user.

[0174] Optionally, in one embodiment of this application, an index is established for at least one patent database and / or at least one thesis database to retrieve patent data and / or thesis data, including: generating at least one navigation area based on at least one patent database and / or at least one thesis database; generating at least one daily dynamic area based on at least one patent database and / or at least one thesis database; generating at least one technical service area based on at least one patent database and / or at least one thesis database; and generating at least one research results area based on at least one patent database and / or at least one thesis database.

[0175] Optionally, in one embodiment of this application, based on patent data and / or paper data, a preset model context protocol is used to perform semantic retrieval on the target database to filter out results that do not meet the multi-dimensional conditions, and obtain retrieval results, including: performing multi-field composite weight query, multi-dimensional filtering and sorting in IPC classification, application year and relevance; performing multi-dimensional retrieval and filtering in title, abstract, journal, citation count; and performing semantic efficient retrieval of the ElasticSearch database based on the model context protocol, supporting multi-dimensional condition filtering and high-concurrency access.

[0176] Optionally, in one embodiment of this application, the search results are cleaned and standardized to obtain final data that meets preset data quality and consistency conditions, and the user's intelligence generation needs are identified according to the analysis type selected by the user, including: conducting analysis on patent technology development trends, inventor and applicant distribution, and technology hotspots; conducting analysis on research topics, author contributions, and citations; conducting analysis on the professional fields, innovation capabilities, and cooperation networks of research institutions; conducting evaluation of cooperation relationships among enterprises, universities, and research institutes; and conducting multi-dimensional comparative analysis of different institutions or patent groups.

[0177] Optionally, in one embodiment of this application, a pre-set large language model is used to generate a corresponding professional analysis report based on the final data and intelligence generation requirements, including: automatically constructing specialized prompt words according to the analysis requirements; integrating multi-source data to form an analysis basis; and automatically generating a structured professional analysis report that supports various report types such as institutional theme trajectory, patents, papers, author academic profiles, industry-academia-research cooperation, regional analysis, and institutional comparison.

[0178] Optionally, in one embodiment of this application, multi-dimensional data visualization information is generated based on a professional analysis report to present the professional analysis report to the user, including: displaying development trends over time; visualizing the relationships and cooperation networks between entities; displaying geographical distribution characteristics and innovation hotspots; and displaying comparative analysis results, including radar charts, heat maps, and timeline charts.

[0179] It should be noted that the foregoing explanation of the embodiment of the intelligent intelligence analysis system based on multimodal technology topic trajectories also applies to the intelligent intelligence analysis method based on multimodal technology topic trajectories in this embodiment, and will not be repeated here.

[0180] The intelligent intelligence analysis method based on multimodal science and technology topic trajectories proposed in this application can establish a database index for retrieval, utilize a preset model context protocol for semantic retrieval and filter results, and after cleaning and standardization, combine user needs to generate professional analysis reports from a large language model and visualize them. This achieves efficient discovery, intelligent analysis, and visualization of science and technology intelligence, effectively improving the efficiency and accuracy of intelligence analysis, allowing users to quickly obtain comprehensive and valuable science and technology intelligence, and providing strong support for decision-making. Therefore, it solves the problems in related technologies that only support a single data source, resulting in the inability to integrate multimodal data for comprehensive analysis, low retrieval efficiency, long response time, inability to meet real-time interaction needs, difficulty in conducting in-depth trend analysis, collaborative analysis, and comparative analysis, and lack of intelligent analysis report generation capabilities.

[0181] Figure 9 A schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device may include:

[0182] The memory 901, the processor 902, and the computer program stored on the memory 901 and capable of running on the processor 902.

[0183] When the processor 902 executes the program, it implements the intelligent intelligence analysis method based on multimodal technology topic trajectories provided in the above embodiments.

[0184] Furthermore, electronic devices also include:

[0185] Communication interface 903 is used for communication between memory 901 and processor 902.

[0186] The memory 901 is used to store computer programs that can run on the processor 902.

[0187] The memory 901 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0188] If the memory 901, processor 902, and communication interface 903 are implemented independently, then the communication interface 903, memory 901, and processor 902 can be interconnected via a bus to complete communication between them. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 9 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0189] Optionally, in a specific implementation, if the memory 901, processor 902, and communication interface 903 are integrated on a single chip, then the memory 901, processor 902, and communication interface 903 can communicate with each other through an internal interface.

[0190] The processor 902 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application.

[0191] This embodiment also provides a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described intelligent intelligence analysis method based on multimodal technology topic trajectories.

[0192] This application also provides a computer program product storing a computer program that, when executed by a processor, implements the above-described intelligent intelligence analysis method based on multimodal technology topic trajectories.

[0193] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0194] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0195] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or N executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0196] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0197] It should be understood that the various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, it can be implemented using any one or more of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0198] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0199] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0200] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. An intelligent intelligence analysis system based on multimodal technology theme trajectories, characterized in that, include: The homepage module is used to create an index to at least one patent database and / or at least one paper database to retrieve patent data and / or paper data; The multimodal data search module is used to perform semantic retrieval of the target database based on the patent data and / or the paper data using a preset model context protocol, so as to filter out the results that do not meet the multi-dimensional conditions and obtain the search results. The intelligent analysis module is used to clean and standardize the search results to obtain final data that meets preset data quality and consistency conditions, and to identify the user's intelligence generation needs based on the analysis type selected by the user. An AI-driven engine is used to generate corresponding professional analysis reports based on the final data and the intelligence generation requirements using a preset large language model. The visualization module is used to generate multi-dimensional data visualization information based on the professional analysis report in order to present the professional analysis report to the user.

2. The system according to claim 1, characterized in that, The homepage module includes: The functional navigation submodule is used to generate at least one navigation area based on the at least one patent database and / or at least one paper database; The user management submodule is used to generate at least one daily dynamic zone based on the at least one patent database and / or at least one paper database. The system configuration submodule is used to generate at least one technical service area based on the at least one patent database and / or at least one thesis database; The Daily Dynamics submodule is used to generate at least one research results area based on the at least one patent database and / or at least one paper database.

3. The system according to claim 1, characterized in that, The multimodal data search module includes: The patent search submodule is used for multi-field composite weighted queries, IPC classification, multi-dimensional filtering and sorting based on application year and relevance; The paper search submodule is used for multi-dimensional searching and filtering based on title, abstract, journal, and citation count. An intelligent data acquisition engine based on the model context protocol is used for semantic and efficient retrieval of ElasticSearch databases based on the model context protocol, supporting multi-dimensional condition filtering and high-concurrency access.

4. The system according to claim 1, characterized in that, The intelligent analysis module includes: The patent analysis submodule is used to analyze patent technology development trends, inventor and applicant distribution, and technological hotspots. The paper analysis submodule is used to analyze research topics, author contributions, and citations. The author analysis submodule is used to analyze the research institution's area of ​​expertise, innovation capabilities, and collaborative networks; The industry-academia-research collaboration analysis submodule is used to evaluate the collaborative relationships between enterprises, universities, and research institutions. The Institutional Comparison Analysis submodule is used to conduct multi-dimensional comparative analysis of different institutions or patent groups.

5. The system according to claim 1, characterized in that, The AI-driven engine includes: The prompt engineering submodule is used to automatically generate specialized prompt words based on analysis requirements; The data fusion submodule is used to integrate multi-source data to form the basis for analysis; The report generation submodule is used to automatically generate structured professional analysis reports, supporting various report types such as institutional theme trajectory, patents, papers, author academic profiles, industry-academia-research collaborations, regional analysis, and institutional comparisons.

6. The system according to claim 1, characterized in that, The visualization module includes: The trend chart submodule is used to display development trends over a time series. The Relationship Network submodule is used to visualize the relationships and cooperation networks between entities; The distribution map submodule is used to display the distribution characteristics and innovation hotspots in geospatial space; The comparison charts submodule is used to display comparative analysis results, including radar charts, heatmaps, and timeline charts.

7. An intelligent intelligence analysis method based on multimodal technology theme trajectories, characterized in that, Includes the following steps: Establish indexes for at least one patent database and / or at least one academic paper database to retrieve patent data and / or academic paper data; Based on the patent data and / or the paper data, a semantic search is performed on the target database using a preset model context protocol to filter out results that do not meet the multi-dimensional conditions and obtain the search results. The search results are cleaned and standardized to obtain final data that meets preset data quality and consistency conditions, and the user's intelligence generation needs are identified according to the analysis type selected by the user. A corresponding professional analysis report is generated based on the final data and the intelligence generation requirements using a pre-defined large language model. Based on the professional analysis report, multi-dimensional data visualization information is generated to present the professional analysis report to the user.

8. An electronic device, characterized in that, include: The memory, the processor, and the computer program stored in the memory and executable on the processor, the processor executing the program to implement the intelligent intelligence analysis method based on multimodal technology theme trajectories as described in claim 7.

9. A non-volatile computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to implement the intelligent intelligence analysis method based on multimodal technology topic trajectories as described in claim 7.

10. A computer program product, comprising a computer program, characterized in that, The computer program is executed to implement the intelligent intelligence analysis method based on multimodal technology topic trajectories as described in claim 7.

Images