Patent risk early warning system driven by artificial intelligence

Abstract

The invention relates to the technical field of patent early warning, and discloses an artificial intelligence-driven patent risk early warning system, and the system comprises a data collection module which is used for collecting patent data from a patent data source, and carrying out the preprocessing and storage of the data; the patent analysis module is used for carrying out information analysis on the patent texts based on natural language processing and machine learning technologies so as to obtain various data of patents and carrying out classification and similarity analysis on the patents; and the risk assessment module is used for assessing risk analysis of a target patent in combination with various data of the patent, analyzing the legal status of the patent, and assessing historical litigation disputes or potential litigation risks of the patent. According to the invention, through cooperative operation of the data acquisition module, the patent analysis module, the risk assessment module, the risk early warning module and the feedback optimization module, advanced natural language processing, machine learning and other technologies are utilized, and the effects of comprehensive and accurate assessment and early warning of patent risks and continuous optimization are achieved.

Description

Technical Field

[0001] The present invention relates to the field of patent early warning technology, and in particular to an artificial intelligence-driven patent risk early warning system. Background Art

[0002] Patent early warning refers to the mechanism of collecting and analyzing patent information and domestic and foreign market information in the technical field of the industry and related technical fields to understand what competitors are doing, and promptly inform relevant government departments, industry organizations and companies in the industry of the precursors of possible patent disputes and possible harms, as well as recommended countermeasures and measures; at the same time, it releases information on infringement of patent rights and recommends that industry organizations and companies in the industry take countermeasures. Patent early warning analysis is mainly divided into national patent early warning analysis, industry patent early warning analysis and enterprise patent early warning analysis.

[0003] The traditional method of patent risk warning is often to organize expert evaluation and demonstration to issue risk warnings for patents, but manual analysis and warnings are often inefficient. Summary of the invention

[0004] In order to make up for the above shortcomings, the present invention provides an artificial intelligence-driven patent risk warning system, which aims to improve the traditional patent risk warning method, which often conducts risk warnings on patents by organizing expert evaluation and demonstration, while manual analysis and warnings are often inefficient.

[0005] In the first aspect, the present invention provides the following technical solution: an artificial intelligence-driven patent risk early warning system, comprising:

[0006] Data collection module: used to collect patent data from patent data sources, and perform pre-processing and storage;

[0007] Patent analysis module: used to analyze patent texts based on natural language processing and machine learning technologies, thereby obtaining various data of patents, and classifying and similarity analyzing patents;

[0008] Risk assessment module: used to combine various data of patents, evaluate the risk analysis of target patents, analyze the legal status of patents, evaluate the historical litigation disputes or potential litigation risks of the patents, evaluate the potential risks brought by the competition situation, quantify patent risks and generate risk scores;

[0009] Risk warning module: used to generate and push warning information based on the assessed risk score and preset thresholds;

[0010] Feedback optimization module: used to collect user feedback and continuously optimize the system based on the feedback.

[0011] Preferably, the data acquisition module includes a data capture unit, a data cleaning unit, and a data storage unit. The data capture unit is used to collect patent data from patent data sources using web crawler technology combined with search algorithms. The patent data sources include patent database websites, government patent announcement platforms, and corporate patent libraries. The patent data include patent names, applicants, invention contents, and claims. The data cleaning unit is used to pre-process the captured original patent data using text pre-processing technology in natural language processing, and to correct the data format according to the standard format requirements of the patent data. The pre-processing includes removing noise and invalid information from the data. The data storage unit is used to store the corrected data in a database, and use database management system technology combined with artificial intelligence indexing and optimization algorithms to store and organize the data.

[0012] Preferably, the patent analysis module includes a text parsing unit, a technical field classification unit, and a similarity analysis unit. The text parsing unit is used to parse patent documents through lexical analysis, syntactic analysis, and semantic analysis methods in natural language processing technology. The parsing is to extract various data of the patent by constructing a language model of the patent field. The various data include invention content, technical features, key terms, technical entities, and relationships. The technical field classification unit is used to classify patents according to the content in the patent text through a classification algorithm in machine learning and classify them into corresponding technical fields. The similarity analysis unit is used to perform similarity analysis on patents through a machine learning algorithm. The similarity analysis is to calculate the similarity between different patents by converting the patent text into a vector representation.

[0013] Preferably, the risk assessment module includes a risk analysis unit and a risk scoring unit. The risk analysis unit is used to perform risk analysis on various data of the patent, wherein the various data include historical data, legal status, technical field and patent similarity. The risk scoring unit is used to generate a comprehensive risk score of the patent based on the data obtained by the risk analysis unit, using weighted summation or hierarchical analysis method in combination with a regression model in machine learning.

[0014] Preferably, the risk analysis unit includes a legal status analysis subunit, a litigation record analysis subunit, and a competitive situation analysis subunit. The legal status analysis subunit is used to analyze the legal status of the patent based on the legal status information of the patent by using a rule engine combined with the knowledge graph technology of artificial intelligence. The legal status includes the validity period, invalidity, and revocation status. The litigation record analysis subunit is used to analyze the relationship between the historical litigation data and the current patent status based on the historical litigation data of the patent through data mining technology and prediction models in machine learning, and evaluate the historical litigation disputes or potential litigation risks of the patent. The historical litigation data includes the causes and results of the litigation. The competitive situation analysis subunit is used to identify the patent clusters and technical advantage areas of competitors by using network analysis technology and clustering algorithms in machine learning, and evaluate the potential risks brought by the competitive situation by comparing the technical features and layout of the target patent with those of the competitor's patent.

[0015] Preferably, the risk warning module includes a risk threshold setting unit, a warning information generating unit, and a notification push unit. The risk threshold setting unit is used to set the risk warning threshold according to corresponding needs by using user interface design and configuration management technology, and optimize the threshold setting by using adaptive algorithms in machine learning based on historical warning data and feedback information. The corresponding needs include specific business scenarios and risk tolerance. The warning information generating unit is used to generate a risk warning report for the patent based on the risk score and threshold by using template generation technology and natural language generation technology, and divide the patent risk into high, medium and low levels according to the risk warning report. The risk warning report includes legal status, patent similarity, and patent validity. The notification push unit is used to push the risk warning report to the user by using an intelligent push algorithm, and adjust the notification method and time according to the user's preferences and usage habits.

[0016] Preferably, the feedback optimization module includes a feedback collection unit, a model updating unit, and a data enhancement unit. The feedback collection unit is used to provide a user interface and provide a feedback entry through the user interface, collect feedback information, and automatically classify the feedback using text classification technology. The information includes warning accuracy. The model updating unit is used to adjust the parameters in the risk assessment model and the machine learning algorithm based on the feedback information using reinforcement learning and incremental learning techniques. The data enhancement unit is used to combine the feedback information and the newly acquired patent information and enhance the data set using data expansion technology.

[0017] In a second aspect, the present invention provides the following technical solution, an artificial intelligence-driven patent risk early warning method, comprising the following steps:

[0018] S1. Data collection: Use web crawler technology combined with search algorithms to collect patent data from patent data sources, and then pre-process the patent data through text pre-processing technology in natural language processing, and modify the data format according to the standard format requirements of patent data, and then store the modified data in the database, and use database management system technology combined with artificial intelligence indexing and optimization algorithms to store and organize the data. The patent data sources include patent database websites, government patent announcement platforms, and corporate patent libraries. The patent data include patent names, applicants, invention contents, and claims. The pre-processing includes removing noise and invalid information from the data;

[0019] S2. Patent analysis: Then, the patent documents are parsed through the lexical analysis, syntactic analysis and semantic analysis methods in natural language processing technology, and the patents are classified into corresponding technical fields through the classification algorithm in machine learning according to the content in the patent text, and then similarity analysis is performed on the patents through the machine learning algorithm. The analysis is to extract various data of the patents by constructing a language model of the patent field, and the various data include invention content, technical features, key terms, technical entities and relationships. The similarity analysis is to convert the patent text into a vector representation and calculate the similarity between different patents;

[0020] S3. Risk assessment: Then, risk analysis is performed on various data of the target patent, and based on the data obtained from the risk analysis, a weighted summation or hierarchical analysis method is used, combined with a regression model in machine learning, to generate a comprehensive risk score for the patent. The various data include historical data, legal status, technical field, and patent similarity. The risk analysis includes analyzing the legal status of the patent, evaluating the historical litigation disputes or potential litigation risks of the patent, and evaluating the potential risks brought by the competitive situation;

[0021] S4. Risk warning: Then, according to the corresponding needs, user interface design and configuration management technology are used to set the threshold of risk warning, and according to the historical warning data and feedback information, the adaptive algorithm in machine learning is used to optimize the threshold setting. Then, based on the risk score and threshold, the template generation technology and natural language generation technology are used to generate a risk warning report for the patent, and according to the risk warning report, the patent risk is divided into high, medium and low levels. At the same time, the intelligent push algorithm is used to push the risk warning report to the user, and the notification method and time are adjusted according to the user's preferences and usage habits. The corresponding needs include specific business scenarios and risk tolerance, and the risk warning report includes legal status, patent similarity, and patent validity;

[0022] S5. Feedback optimization: Feedback information is then collected through the user interface and automatically classified using text classification technology. Based on the feedback information, reinforcement learning and incremental learning techniques are used to adjust the parameters in the risk assessment model and machine learning algorithm. The feedback information and newly acquired patent information are combined to enhance the data set using data augmentation technology, and the information includes warning accuracy.

[0023] In the third aspect, the invention provides the following technical solution: a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned artificial intelligence-driven patent risk warning method when executing the computer program.

[0024] In a fourth aspect, the present invention provides the following technical solution: a readable storage medium having a computer program stored thereon, and the computer program, when executed by a processor, implements the above-mentioned artificial intelligence-driven patent risk warning method.

[0025] The present invention has the following beneficial effects:

[0026] 1. In the present invention, through the coordinated operation of the data collection module, the patent analysis module, the risk assessment module, the risk warning module and the feedback optimization module, the advanced natural language processing, machine learning and other technologies are used to achieve a comprehensive and accurate assessment and warning of patent risks, and the effect of continuous optimization can be achieved, which effectively helps enterprises and related institutions to protect patent rights and interests, reasonably respond to competition, and improve the level of patent management. It also improves the traditional patent risk warning method, which is often to conduct risk warning of patents by organizing expert evaluation and demonstration, while manual analysis and warning are often inefficient.

[0027] 2. In the present invention, by using artificial intelligence technology and a series of steps including data collection, patent analysis, risk assessment, risk warning, and feedback optimization, the effect of orderly processing patent data, accurately assessing risks, timely warning, and improving system performance based on feedback is achieved. The problems of messy data, insufficient analysis depth, vague risk judgment, poor quality of warning information, and inability of the system to improve itself in the patent risk warning process are solved, thereby ensuring the scientificity and practicality of patent risk warning and assisting users to better manage patent risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 A system architecture diagram of the artificial intelligence-driven patent risk early warning system proposed in the present invention;

[0029] Figure 2 This is a data collection module architecture diagram of the artificial intelligence-driven patent risk early warning system proposed in the present invention;

[0030] Figure 3 The patent analysis module architecture diagram of the artificial intelligence-driven patent risk early warning system proposed in the present invention;

[0031] Figure 4 The risk assessment module architecture diagram of the artificial intelligence-driven patent risk early warning system proposed in the present invention;

[0032] Figure 5 The risk warning module architecture diagram of the artificial intelligence-driven patent risk warning system proposed in the present invention;

[0033] Figure 6 This is a feedback optimization module architecture diagram of the artificial intelligence-driven patent risk early warning system proposed in the present invention;

[0034] Figure 7 A method flow chart of the artificial intelligence-driven patent risk early warning method proposed in the present invention. DETAILED DESCRIPTION

[0035] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0036] Embodiment 1

[0037] Reference Figure 1 In a first embodiment of the present invention, the present invention provides an artificial intelligence driven patent risk early warning system, comprising:

[0038] Data collection module: used to collect patent data from patent data sources, and perform pre-processing and storage;

[0039] Patent analysis module: used to analyze patent texts based on natural language processing and machine learning technologies, thereby obtaining various data of patents, and classifying and similarity analyzing patents;

[0040] Risk assessment module: used to combine various data of patents, evaluate the risk analysis of target patents, analyze the legal status of patents, evaluate the historical litigation disputes or potential litigation risks of the patents, evaluate the potential risks brought by the competition situation, quantify patent risks and generate risk scores;

[0041] Risk warning module: used to generate and push warning information based on the assessed risk score and preset thresholds;

[0042] Feedback optimization module: used to collect user feedback and continuously optimize the system based on the feedback.

[0043] Specifically, the data acquisition module collects patent data from patent data sources, and performs preprocessing and storage, thereby providing a sufficient basis for subsequent analysis, making risk warning more comprehensive and accurate, improving data quality, ensuring the accuracy and reliability of subsequent analysis, improving data management efficiency, and accelerating the response speed of the risk warning system;

[0044] The patent analysis module uses natural language processing and machine learning technology to analyze patent texts, thereby obtaining various data on patents, and classifying and similarity analyzing patents, so as to gain a deeper understanding of patents, help accurately assess risks, understand patent distribution, provide references for enterprises to formulate technology strategies and risk prevention, discover potential infringement risks, and help enterprises protect their own patents and avoid infringement disputes;

[0045] The risk assessment module combines various data of patents to evaluate the risk analysis of target patents, analyze the legal status of patents, evaluate the historical litigation disputes or potential litigation risks of the patents, and evaluate the potential risks brought by the competitive situation. It also quantifies patent risks and generates risk scores, thereby providing a comprehensive basis for comprehensive evaluation, allowing enterprises to know risks in advance, formulate response strategies, and intuitively present the level of risks, so that enterprises can quickly judge the degree of risk and determine the priority of response measures;

[0046] The risk warning module generates and pushes warning information based on the assessed risk scores and preset thresholds, thereby reducing false positives and missed positives, making warnings more realistic, avoiding waste of enterprise resources and missing important risks, and providing users with clear and comprehensive risk warnings to help them accurately grasp the risk status and take effective measures to ensure that users receive information in a timely manner, so as to quickly respond to risks and reduce losses.

[0047] Collect user feedback through the feedback optimization module, and continuously optimize the system based on the feedback, so as to understand user needs and system deficiencies, provide direction for optimization, make the system more in line with user expectations, improve the accuracy of risk assessment, adapt the system to new situations, improve the reliability of early warning, enrich data resources, further improve system performance and analysis results, and make risk early warning more accurate and effective;

[0048] Through the coordinated operation of various modules of this system, advanced natural language processing, machine learning and other technologies are utilized to achieve a comprehensive and accurate assessment and early warning of patent risks, and the effect of continuous optimization is achieved. This improves the traditional patent risk early warning method, which often involves organizing expert evaluation and demonstration to issue risk warnings for patents, while manual analysis and early warnings are often inefficient. This effectively helps companies and related institutions protect patent rights, reasonably respond to competition, and improve patent management levels.

[0049] Reference Figure 2The data acquisition module includes a data capture unit, a data cleaning unit, and a data storage unit. The data capture unit is used to collect patent data from patent data sources using web crawler technology combined with search algorithms. Patent data sources include patent database websites, government patent announcement platforms, and corporate patent libraries. Patent data includes patent names, applicants, invention contents, and claims. The data cleaning unit is used to pre-process the captured original patent data through text pre-processing technology in natural language processing, and correct the data format according to the standard format requirements of patent data. Pre-processing includes removing noise and invalid information from the data. The data storage unit is used to store the corrected data in the database, and use database management system technology combined with artificial intelligence indexing and optimization algorithms to store and organize the data.

[0050] Specifically, the data capture unit collects patent data from multiple data sources to ensure rich data, provide a sufficient basis for subsequent analysis, and make risk warnings more comprehensive and accurate. The data cleaning unit preprocesses data to remove noise, corrects the format, standardizes data, improves data quality, and ensures the accuracy and reliability of subsequent analysis. The data storage unit effectively stores and organizes data to facilitate rapid retrieval and utilization, improves data management efficiency, and accelerates the response speed of the risk warning system.

[0051] Reference Figure 3 The patent analysis module includes a text parsing unit, a technical field classification unit, and a similarity analysis unit. The text parsing unit is used to parse patent documents through lexical analysis, syntactic analysis, and semantic analysis methods in natural language processing technology. The parsing is to extract various data of the patent by building a language model in the patent field. The various data include invention content, technical features, key terms, technical entities, and relationships. The technical field classification unit is used to classify patents according to the content in the patent text through the classification algorithm in machine learning and classify them into corresponding technical fields. The similarity analysis unit is used to perform similarity analysis on patents through machine learning algorithms. Similarity analysis is to convert patent texts into vector representations and calculate the similarity between different patents.

[0052] Specifically, the text parsing unit deeply mines patent text information to provide accurate data for classification and similarity analysis, which enables a deeper understanding of patents and helps to accurately assess risks. The technical field classification unit accurately classifies patents into corresponding fields, which is conducive to understanding patent distribution and providing reference for enterprises to formulate technical strategies and risk prevention. The similarity analysis unit calculates patent similarity, which can discover potential infringement risks, help enterprises protect their own patents, and avoid infringement disputes.

[0053] Reference Figure 4The risk assessment module includes a risk analysis unit and a risk scoring unit. The risk analysis unit is used to perform risk analysis on various data of the patent, including historical data, legal status, technical field and patent similarity. The risk scoring unit is used to generate a comprehensive risk score of the patent based on the data obtained by the risk analysis unit, using weighted summation or hierarchical analysis method, combined with the regression model in machine learning; the risk analysis unit includes a legal status analysis subunit, a litigation record analysis subunit, and a competitive situation analysis subunit. The legal status analysis subunit is used to analyze the legal status of the patent based on the legal status information of the patent using a rule engine combined with the knowledge graph technology of artificial intelligence. The legal status includes the validity period, invalidation, and revocation status. The litigation record analysis subunit is used to analyze the relationship between the historical litigation data and the current patent status based on the historical litigation data of the patent through data mining technology and the prediction model in machine learning, and evaluate the historical litigation disputes or potential litigation risks of the patent. The historical litigation data includes the causes and results of the litigation. The competitive situation analysis subunit is used to use network analysis technology and clustering algorithms in machine learning to identify the patent clusters and technical advantages of competitors, and evaluate the potential risks brought by the competitive situation by comparing the technical features and layout of the target patent with those of the competitor's patent.

[0054] Specifically, each sub-unit of the risk analysis unit accurately analyzes patent law, litigation, and competition risks, providing a comprehensive basis for comprehensive evaluation, enabling enterprises to know the risks in advance and formulate response strategies. The risk scoring unit quantifies the risks and generates scores, which intuitively present the level of risk, making it easier for enterprises to quickly judge the degree of risk and determine the priority of response measures.

[0055] Reference Figure 5 The risk warning module includes a risk threshold setting unit, a warning information generation unit, and a notification push unit. The risk threshold setting unit is used to set the risk warning threshold according to the corresponding needs, using the user interface design and configuration management technology, and optimize the threshold setting based on the historical warning data and feedback information using the adaptive algorithm in machine learning. The corresponding needs include specific business scenarios and risk tolerance. The warning information generation unit is used to generate a patent risk warning report based on the risk score and threshold, using template generation technology and natural language generation technology, and divide the patent risk into high, medium and low levels according to the risk warning report. The risk warning report includes legal status, patent similarity, and patent validity. The notification push unit is used to push the risk warning report to the user using the intelligent push algorithm, and adjust the notification method and time according to the user's preferences and usage habits.

[0056] Specifically, the risk threshold setting unit optimizes the threshold according to needs and data, reduces false alarms and missed alarms, makes warnings more realistic, avoids wasting enterprise resources and missing important risks; the warning information generation unit generates detailed reports and grades them to provide users with clear and comprehensive risk warnings, helping them to accurately grasp the risk status and take effective measures; the notification push unit pushes reports in a timely manner according to user preferences, ensuring that users receive information in a timely manner, facilitating rapid response to risks and reducing losses.

[0057] Reference Figure 6 ,The feedback optimization module includes a feedback collection unit, a model updating unit, and a data ,enhancement unit. The feedback collection unit is used to provide a user interface and a feedback ,entry through the user interface, collect feedback information, and automatically ,classify the feedback using text classification technology. The information includes warning accuracy. The model updating unit is used to ,adjust the parameters in the risk assessment model and the machine learning ,algorithm based on the feedback information using reinforcement learning and incremental ,learning techniques. The data enhancement unit is used to combine the feedback ,information and the newly acquired patent information and enhance the ,dataset using data expansion technology.

[0058] Specifically, the feedback collection unit collects classified feedback to understand user needs and system deficiencies, provide direction for optimization, and make the system more in line with user expectations. The model updating unit adjusts model parameters according to feedback to improve the accuracy of risk assessment, adapt the system to new situations, and improve the reliability of early warning. The data set is expanded and data resources are enriched through the data enhancement unit to further improve system performance and analysis results, making risk warnings more accurate and effective.

[0059] Embodiment 2:

[0060] Reference Figure 7 In a second embodiment of the present invention, the present invention provides an artificial intelligence-driven patent risk early warning method, comprising the following steps:

[0061] S1. Data collection: Use web crawler technology combined with search algorithms to collect patent data from patent data sources, and then pre-process the patent data through text pre-processing technology in natural language processing, and modify the data format according to the standard format requirements of patent data, and then store the modified data in the database, and use database management system technology combined with artificial intelligence indexing and optimization algorithms to store and organize the data. Patent data sources include patent database websites, government patent announcement platforms, and corporate patent libraries. Patent data include patent names, applicants, invention contents, and claims. Pre-processing includes removing noise and invalid information in the data;

[0062] S2. Patent analysis: Then, the patent documents are parsed through the lexical analysis, syntactic analysis and semantic analysis methods in natural language processing technology, and the patents are classified into the corresponding technical fields through the classification algorithm in machine learning according to the content in the patent text. Then, the patents are analyzed for similarity through the machine learning algorithm. The analysis is to extract various data of the patent by building a language model in the patent field. The data includes invention content, technical features, key terms, technical entities and relationships. The similarity analysis is to convert the patent text into a vector representation and calculate the similarity between different patents.

[0063] S3. Risk assessment: Then, we conduct risk analysis on various data of the target patent, and generate a comprehensive risk score of the patent based on the data obtained from the risk analysis, using weighted summation or hierarchical analysis combined with the regression model in machine learning. The data includes historical data, legal status, technical field and patent similarity. The risk analysis includes analyzing the legal status of the patent, evaluating the historical litigation disputes or potential litigation risks of the patent, and evaluating the potential risks brought by the competitive situation.

[0064] S4. Risk warning: Then, according to the corresponding needs, user interface design and configuration management technology are used to set the threshold of risk warning, and according to the historical warning data and feedback information, the adaptive algorithm in machine learning is used to optimize the threshold setting. Then, based on the risk score and threshold, the template generation technology and natural language generation technology are used to generate a risk warning report for the patent. According to the risk warning report, the patent risk is divided into high, medium and low levels. At the same time, the intelligent push algorithm is used to push the risk warning report to the user, and the notification method and time are adjusted according to the user's preferences and usage habits. The corresponding needs include specific business scenarios and risk tolerance. The risk warning report includes legal status, patent similarity, and patent validity.

[0065] S5. Feedback optimization: Feedback information is then collected through the user interface and automatically classified using text classification technology. Based on the feedback information, reinforcement learning and incremental learning techniques are used to adjust the parameters in the risk assessment model and machine learning algorithm. The feedback information and newly acquired patent information are combined, and data augmentation technology is used to enhance the data set, including warning accuracy.

[0066] Specifically, through the coordinated execution of this method, the use of artificial intelligence technology and the process of this method has achieved the effect of orderly processing of patent data, accurate risk assessment, timely warning and improving system performance based on feedback, solving the problems of messy data, insufficient analysis depth, vague risk judgment, poor quality of warning information and inability of the system to improve itself in the patent risk warning process, ensuring the scientificity and practicality of patent risk warning, and assisting users to better manage patent risks.

[0067] Embodiment 3

[0068] The third embodiment of the present invention is based on the same inventive concept. The present invention proposes a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the steps of the artificial intelligence-driven patent risk warning method of the above embodiment.

[0069] Embodiment 4

[0070] The fourth embodiment of the present invention is based on the same inventive concept. The present invention proposes a computer device, including: a processor and a memory; the processor and the memory communicate with each other; the memory is used to store instructions; the processor is used to execute the instructions in the memory, and execute the artificial intelligence driven patent risk warning method of the above embodiment.

[0071] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0072] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. The patent risk early warning system driven by artificial intelligence is characterized by: include: Data collection module: used to collect patent data from patent data sources, and perform pre-processing and storage; Patent analysis module: used to analyze patent texts based on natural language processing and machine learning technologies, thereby obtaining various data of patents, and classifying and similarity analyzing patents; Risk assessment module: used to combine various data of patents, evaluate the risk analysis of target patents, analyze the legal status of patents, evaluate the historical litigation disputes or potential litigation risks of the patents, evaluate the potential risks brought by the competition situation, quantify patent risks and generate risk scores; Risk warning module: used to generate and push warning information based on the assessed risk score and preset thresholds; Feedback optimization module: used to collect user feedback and continuously optimize the system based on the feedback.

2. The artificial intelligence-driven patent risk early warning system according to claim 1, characterized in that: The data acquisition module includes a data capture unit, a data cleaning unit, and a data storage unit. The data capture unit is used to collect patent data from patent data sources using web crawler technology combined with search algorithms. The patent data sources include patent database websites, government patent announcement platforms, and corporate patent libraries. The patent data includes patent names, applicants, invention contents, and claims. The data cleaning unit is used to pre-process the captured original patent data using text pre-processing technology in natural language processing, and to correct the data format according to the standard format requirements of the patent data. The pre-processing includes removing noise and invalid information from the data. The data storage unit is used to store the corrected data in a database, and use database management system technology combined with artificial intelligence indexing and optimization algorithms to store and organize the data.

3. The artificial intelligence-driven patent risk early warning system according to claim 1, characterized in that: The patent analysis module includes a text parsing unit, a technical field classification unit, and a similarity analysis unit. The text parsing unit is used to parse patent documents through lexical analysis, syntactic analysis, and semantic analysis methods in natural language processing technology. The parsing is to extract various data of the patent by constructing a language model of the patent field. The various data include invention content, technical features, key terms, technical entities, and relationships. The technical field classification unit is used to classify patents according to the content in the patent text through a classification algorithm in machine learning and classify them into corresponding technical fields. The similarity analysis unit is used to perform similarity analysis on patents through a machine learning algorithm. The similarity analysis is to calculate the similarity between different patents by converting the patent text into a vector representation.

4. The artificial intelligence-driven patent risk early warning system according to claim 1, characterized in that: The risk assessment module includes a risk analysis unit and a risk scoring unit. The risk analysis unit is used to perform risk analysis on various data of the patent, including historical data, legal status, technical field and patent similarity. The risk scoring unit is used to generate a comprehensive risk score of the patent based on the data obtained by the risk analysis unit, using weighted summation or hierarchical analysis method, combined with a regression model in machine learning.

5. The artificial intelligence-driven patent risk early warning system according to claim 4 is characterized in that: The risk analysis unit includes a legal status analysis subunit, a litigation record analysis subunit, and a competitive situation analysis subunit. The legal status analysis subunit is used to analyze the legal status of the patent based on the legal status information of the patent by using a rule engine combined with the knowledge graph technology of artificial intelligence. The legal status includes validity period, invalidity, and revocation status. The litigation record analysis subunit is used to analyze the relationship between historical litigation data and current patent status based on the historical litigation data of the patent through data mining technology and prediction models in machine learning, and evaluate the historical litigation disputes or potential litigation risks of the patent. The historical litigation data includes the causes and results of the litigation. The competitive situation analysis subunit is used to identify the patent clusters and technical advantage areas of competitors by using network analysis technology and clustering algorithms in machine learning, and evaluate the potential risks brought by the competitive situation by comparing the technical features and layout of the target patent with those of the competitor's patent.

6. The artificial intelligence-driven patent risk early warning system according to claim 1, characterized in that: The risk warning module includes a risk threshold setting unit, a warning information generation unit, and a notification push unit. The risk threshold setting unit is used to set the risk warning threshold according to corresponding needs by using user interface design and configuration management technology, and optimize the threshold setting by using adaptive algorithms in machine learning based on historical warning data and feedback information. The corresponding needs include specific business scenarios and risk tolerance. The warning information generation unit is used to generate a risk warning report for the patent based on risk scores and thresholds by using template generation technology and natural language generation technology, and divide the patent risk into high, medium and low levels according to the risk warning report. The risk warning report includes legal status, patent similarity, and patent validity. The notification push unit is used to push the risk warning report to the user by using an intelligent push algorithm, and adjust the notification method and time according to the user's preferences and usage habits.

7. The artificial intelligence-driven patent risk early warning system according to claim 1, characterized in that: The feedback optimization module includes a feedback collection unit, a model updating unit, and a data enhancement unit. The feedback collection unit is used to provide a user interface and provide a feedback entry through the user interface, collect feedback information, and automatically classify the feedback using text classification technology. The information includes warning accuracy. The model updating unit is used to adjust the parameters in the risk assessment model and the machine learning algorithm based on the feedback information using reinforcement learning and incremental learning techniques. The data enhancement unit is used to combine the feedback information and the newly acquired patent information and enhance the data set using data expansion technology.

8. The patent risk early warning method driven by artificial intelligence is characterized by: The artificial intelligence-driven patent risk early warning system applied to any one of claims 1 to 7 comprises the following steps: S1. Data collection: Use web crawler technology combined with search algorithms to collect patent data from patent data sources, and then pre-process the patent data through text pre-processing technology in natural language processing, and modify the data format according to the standard format requirements of patent data, and then store the modified data in the database, and use database management system technology combined with artificial intelligence indexing and optimization algorithms to store and organize the data. The patent data sources include patent database websites, government patent announcement platforms, and corporate patent libraries. The patent data include patent names, applicants, invention contents, and claims. The pre-processing includes removing noise and invalid information from the data; S2. Patent analysis: Then, the patent documents are parsed through the lexical analysis, syntactic analysis and semantic analysis methods in natural language processing technology, and the patents are classified into corresponding technical fields through the classification algorithm in machine learning according to the content in the patent text, and then similarity analysis is performed on the patents through the machine learning algorithm. The analysis is to extract various data of the patents by constructing a language model of the patent field, and the various data include invention content, technical features, key terms, technical entities and relationships. The similarity analysis is to convert the patent text into a vector representation and calculate the similarity between different patents; S3. Risk assessment: Then, risk analysis is performed on various data of the target patent, and based on the data obtained from the risk analysis, a weighted summation or hierarchical analysis method is used, combined with a regression model in machine learning, to generate a comprehensive risk score for the patent. The various data include historical data, legal status, technical field, and patent similarity. The risk analysis includes analyzing the legal status of the patent, evaluating the historical litigation disputes or potential litigation risks of the patent, and evaluating the potential risks brought by the competitive situation; S4. Risk warning: Then, according to the corresponding needs, user interface design and configuration management technology are used to set the threshold of risk warning, and according to the historical warning data and feedback information, the adaptive algorithm in machine learning is used to optimize the threshold setting. Then, based on the risk score and threshold, the template generation technology and natural language generation technology are used to generate a risk warning report for the patent, and according to the risk warning report, the patent risk is divided into high, medium and low levels. At the same time, the intelligent push algorithm is used to push the risk warning report to the user, and the notification method and time are adjusted according to the user's preferences and usage habits. The corresponding needs include specific business scenarios and risk tolerance, and the risk warning report includes legal status, patent similarity, and patent validity; S5. Feedback optimization: Feedback information is then collected through the user interface and automatically classified using text classification technology. Based on the feedback information, reinforcement learning and incremental learning techniques are used to adjust the parameters in the risk assessment model and machine learning algorithm. The feedback information and newly acquired patent information are combined to enhance the data set using data augmentation technology, and the information includes warning accuracy.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the artificial intelligence-driven patent risk warning method as described in claim 8 is implemented.

10. A readable storage medium, characterized in that: The readable storage medium stores a computer program, which, when executed by a processor, implements the artificial intelligence-driven patent risk warning method as described in claim 8.

Images