OA reply method based on intellectual property application process management system AI agent

Abstract

This invention relates to the field of intellectual property service technology, and provides an OA (Office Automation) response method based on an AI agent in an intellectual property application process management system. The method includes: creating an AI agent question-and-answer module and a knowledge base for storing OA document process questions and answers within the intellectual property application process management system; configuring the AI ​​agent question-and-answer module to access the knowledge base via a large language model; obtaining OA documents through the AI ​​agent question-and-answer module, and automatically identifying the OA documents using the large language model to extract preset key features, then retrieving matching opinion processing flows and response strategies related to the key features from the knowledge base; automatically generating OA response content based on the matched opinion processing flows and response strategies using the large language model, and providing the OA response content to the system administrator through the AI ​​agent question-and-answer module. This invention can automate the OA response process, improving the efficiency and accuracy of OA document processing and responses.

Description

Technical Field

[0001] This invention relates to the technical fields of intellectual property services, and in particular to an OA response method based on an AI agent of an intellectual property application process management system. Background Technology

[0002] In transnational intellectual property application practice, especially patent applications, complex interactions with official examination authorities are involved. Receiving and responding to Official Examination Opinions (OAs) is a core and time-consuming step. Existing technologies have attempted to apply artificial intelligence to patent task processing. For example, Patent Document 1 (CN119003124A) discloses a task processing method based on a large model, capable of identifying user intent and invoking patent service tools to handle patent-related tasks such as examination opinions. This method improves task processing efficiency to some extent. Patent Document 2 (CN119850146A) discloses a more general task processing method, which uses a large language model to analyze and break down tasks based on intent, and schedules corresponding business and technical agents for hierarchical execution. However, existing solutions still have significant shortcomings. First, existing methods are mostly passive, reactive task processing. System administrators need to manually download OA files from the official server and then import them into the system for analysis. From receiving the OA notification to starting processing, multiple manual steps are typically required, such as: logging into email / official system, downloading the OA document, saving it locally, opening and reading the document, switching to the analysis system, manually entering questions, checking the processing flow, writing a response outline, and sending feedback to the applicant—all of which are inefficient. Secondly, for multilingual OA documents (such as English, Japanese, and Korean OA documents for different national phases of a PCT application), the existing system cannot automatically recognize and understand them, still requiring professional translation and interpretation, which places extremely high demands on the user's language skills and professional experience.

[0003] Therefore, there is an urgent need for a method that can be deeply integrated into the intellectual property application process to achieve full automation from automatic acquisition and intelligent understanding of OA documents to automatic generation of OA response content, in order to simplify manual operation steps and lower the professional threshold. Summary of the Invention

[0004] To address the shortcomings of the existing technologies, this invention provides an OA response method based on an AI agent in an intellectual property application process management system, which solves the problems of cumbersome OA response process operations, heavy reliance on human understanding and processing, and low system intelligence level.

[0005] In a first aspect, the OA response method based on an AI agent of an intellectual property application process management system provided by the present invention includes: In the intellectual property application process management system, an AI intelligent agent question-answering module and a knowledge base for storing OA document process questions and answers are created. The AI ​​intelligent agent question-answering module is configured so that it can call a large language model to access the knowledge base. The AI ​​intelligent agent question-answering module obtains the OA document and calls the large language model to automatically identify the OA document, extract preset key features, and retrieve the opinion processing flow and response strategy related to the key features from the knowledge base; Based on the matched opinion processing flow and response strategy, the system automatically generates response OA content through the large language model, and provides the response OA content to the internal administrator of the intellectual property application process management system through the AI ​​intelligent agent question-and-answer module.

[0006] Secondly, the aforementioned intellectual property application process management system supports the dynamic generation of patent application forms from multiple countries, specifically including the following steps: Analyze the form content of patent application forms in different target countries, break down the form content into different reusable logical units, set corresponding standardized controls for each logical unit and integrate them into the editor; The editor combines different logical unit controls into patent application form templates that adapt to the requirements of different target countries, and generates corresponding template configuration data based on the patent application form templates. When the patent application form template of a specific target country is accessed, a layered information collection interface adapted to the input habits of Chinese users is generated according to the template configuration data corresponding to the patent application form template. The layered information collection interface contains two independent collection sub-units, namely the applicant and inventor information collection unit and the invention information collection unit, which are used to collect basic patent application data in Chinese format. The basic data of Chinese format patent applications collected by the applicant and inventor information collection unit and the invention information collection unit are configured according to the template to automatically generate a patent application form that conforms to the official specifications of the corresponding target country.

[0007] Compared with the prior art, the beneficial effects of this invention are as follows: This invention provides an OA (Office Automation) response method based on an AI agent within an intellectual property application process management system. The method includes: creating an AI agent question-and-answer module and a knowledge base for storing OA document process questions and answers within the intellectual property application process management system; configuring the AI ​​agent question-and-answer module to access the knowledge base via a large language model; obtaining OA documents through the AI ​​agent question-and-answer module and automatically identifying the OA documents using the large language model, extracting preset key features, and retrieving matching opinion processing flows and response strategies related to the key features from the knowledge base; automatically generating OA response content based on the matched opinion processing flows and response strategies using the large language model, and providing the OA response content to the internal administrators of the intellectual property application process management system through the AI ​​agent question-and-answer module. This invention shortens the OA response coordination cycle and reduces labor costs by automatically communicating with the official server to obtain OA documents and automatically generating OA response content. Furthermore, the large language model automatically identifies and understands multilingual OA documents, eliminating language barriers and automatically generating OA response content, providing professional guidance to system administrators and improving the efficiency and quality of communication with applicants. Attached Figure Description

[0008] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. Some specific embodiments of the invention will be described in detail below with reference to the accompanying drawings in an exemplary and non-limiting manner. The same reference numerals in the drawings designate the same or similar parts or components. It should be understood by those skilled in the art that these drawings are not necessarily drawn to scale. In the drawings: Figure 1 This is a flowchart illustrating an OA response method based on an AI agent in an intellectual property application process management system, according to an embodiment of the present invention. Figure 2 This is a schematic diagram illustrating a process for generating multi-country patent application forms according to an embodiment of the present invention. Detailed Implementation

[0009] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are merely some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention. Example

[0010] See Figures 1-2This embodiment provides an OA (Office Automation) response method based on an AI agent of an intellectual property application process management system, including the following steps: S101. Create an AI intelligent agent question-and-answer module and a knowledge base for storing OA document process questions and answers in the intellectual property application process management system, and configure the AI ​​intelligent agent question-and-answer module so that the AI ​​intelligent agent question-and-answer module can call the large language model to access the knowledge base; S102. Obtain OA documents through the AI ​​intelligent agent question answering module, and call the large language model to automatically identify the OA documents, extract preset key features, and retrieve opinion processing procedures and response strategies related to the key features from the knowledge base; S103. Based on the matched opinion processing flow and response strategy, the response OA content is automatically generated through the large language model, and the response OA content is provided to the internal administrator of the intellectual property application process management system through the AI ​​intelligent agent question and answer module.

[0011] Those skilled in the art will understand that the method provided in this embodiment can be implemented by computer program instructions, which can be stored in a computer-readable storage medium and executed by a computer device. The intellectual property application process management system serves as the basic platform, managing the overall application process. An AI agent question-answering module is embedded in the intellectual property application process management system, preferably built on the Tencent AI agent platform, and can access the knowledge base via the API of a large language model. The communication interface of the intellectual property application process management system can automatically exchange data with the official server, for example, through API interface technology. The knowledge base is used to store intellectual property process question-answer pairs and related processing flows. This knowledge base is preferably built on the Tencent AI agent platform and uses a vector database to store vectorized knowledge. The large language model is used for natural language understanding, intent recognition, text generation, and vectorization training, and can be models with powerful text understanding and generation capabilities, such as the GPT series, Wenxin Yiyan, and Tongyi Qianwen. The user interface of the intellectual property application process management system is preferably a mini-program interface, but can also be an APP interface or a webpage interface, providing an AI agent chat function to support user interaction with the AI ​​agent.

[0012] In this embodiment, an AI agent question-answering module can be developed and integrated into an existing intellectual property application process management system. The construction of this AI agent question-answering module can be quickly achieved by calling API services such as those of the Tencent AI Agent platform. Simultaneously, a knowledge base specifically for storing OA document processing workflows and response strategies is established. The AI ​​agent question-answering module is configured to be able to call a selected large language model (such as GPT-4) via API and has access to the aforementioned knowledge base. Once configured, the AI ​​agent question-answering module begins operation. Its built-in communication interface automatically logs into the designated official server according to preset scheduled tasks (such as 2 AM daily) or trigger conditions (such as receiving a push notification from the official system), retrieves and downloads new OA documents belonging to cases managed by this system. After obtaining the OA document (whether PDF or other format), the AI ​​agent question-answering module calls the large language model to parse it. The large language model can recognize the text content in documents, including multiple languages ​​such as Chinese, English, Japanese, and Korean, and extract pre-defined key features, such as: "application number," "type of examination action notification (e.g., first examination action notification)" explicitly stated in the OA document, "cited legal provisions (e.g., Article 22, Paragraph 3 of the Patent Law)", "comparison of distinguishing technical features pointed out by the examiner", and "reasons for judging obviousness of inventiveness". After extracting key features, these features are converted into query vectors, and similarity searches are performed in a vector knowledge base to match "opinion processing flow" and "response strategy" that have historically handled similar combinations of key features. Based on the matched processing flow and strategy, combined with the specific content of the current OA document, the large language model automatically generates structured response content for the OA. For example, the model can generate: "In response to the examiner's question regarding inventiveness, the following points of argument are suggested: 1. Prior art document 1 does not disclose feature A; 2. The combination of feature A and feature B produces an unexpected technical effect...". The generated content is presented clearly and in a structured manner to the system's internal administrators (such as patent engineers or agents) through the user interface of the AI ​​intelligent agent question-and-answer module, for their reference, modification, or adjustment.

[0013] Preferably, a knowledge base is created for storing OA document process questions and answers, including: Collect the question and answer content of OA documents in the field of intellectual property, extract question and answer features, and form structured question and answer pairs; The text content of the question-and-answer pairs is imported into the knowledge base. The large language model is called, and the vector model is used to transform the text content of the question-and-answer pairs imported into the knowledge base. The text of the question-and-answer pairs is converted into vector representations and stored in the knowledge base to obtain a knowledge base for storing OA document process questions and answers.

[0014] It should be noted that during the system initialization phase, professionals can collect a large amount of historical OA documents and their corresponding successful response cases, relevant provisions of the review guidelines, frequently asked questions, and other textual materials. These materials are processed to extract the "questions" (e.g., "How should I respond when the review opinion cites prior art documents 1 and 2 to evaluate inventiveness?") and "answers" (e.g., the corresponding response strategies and processing procedures), forming structured question-and-answer pairs. The text content of these question-and-answer pairs is then imported in batches into a knowledge base (such as the knowledge base function of Tencent's Intelligent Agent platform). Then, the embedding interface of a large language model is called, and a specific vector model is used to transform the text of each question-and-answer pair, converting the semantic text into high-dimensional vectors. Finally, these high-dimensional vectors and their corresponding original texts are stored in a vector database, completing the knowledge base initialization. This process in this embodiment enables the knowledge base to support semantic similarity-based retrieval, not just keyword matching.

[0015] Preferably, obtaining OA documents through the AI ​​agent question-and-answer module includes: automatically communicating with the official server for intellectual property application management according to preset rules through the communication interface of the AI ​​agent question-and-answer module to obtain new OA documents periodically or in real time. The user interface of the AI ​​agent question-and-answer module is a mini-program interface, APP interface, or web page interface that provides agent chat functionality.

[0016] It should be noted that the communication interface of the AI ​​intelligent agent question-answering module can be programmed to communicate with a specific official server, simulate user login, and periodically (e.g., every hour) check for new OA document lists and download them automatically.

[0017] Preferably, the large language model automatically identifies and understands OA text content in multiple languages, including Chinese. In implementation, a large language model supporting multilingual understanding and generation (such as GPT-4, Claude, etc.) can be selected as the core engine of the system. When the OA document is in English, the model directly reads and understands the English text; when the OA document is in Japanese or Korean, the model, leveraging its multilingual capabilities learned during training, can similarly parse the Japanese or Korean text in the document and extract key information without prior translation, thereby reducing the language proficiency requirements for system users.

[0018] Preferably, the key features include application number, type of examination opinion, legal clauses, comparison of technical features, obviousness assessment, and conventional technical assessment.

[0019] It should be noted that during implementation, prompt word engineering can guide the large language model in feature extraction. Instructions to the large language model could be: "Please extract the following key information from this OA document: application number, type of examination opinion (e.g., novelty, inventiveness, unclear, etc.), cited legal provisions, a detailed description of the examiner's technical feature comparison, the examiner's reasons for the obviousness judgment, and any routine technical judgments (e.g., common knowledge)." Based on these instructions, the large language model locates and structures these feature items in the OA text, providing precise query input for subsequent knowledge base retrieval.

[0020] Preferably, the OA response content provided by the AI ​​intelligent agent question-and-answer module to the internal administrator is presented in a structured form.

[0021] It should be noted that during implementation, the AI ​​agent question-answering module formats the text when returning the OA (Office Automation) response content generated by the large language model to the user interface. For example, Markdown syntax can be used to set different headings such as "Response Ideas," "Suggested Modifications," and "Key Points of Debate," and specific suggestion items can be set as lists, making the final text well-organized and highlighting key points, facilitating quick reading and adoption by administrators.

[0022] Preferably, if no opinion processing flow and response strategy related to the key features are matched in the knowledge base, the content of the OA document and the questions obtained this time are automatically recorded into the knowledge base; the questions recorded into the knowledge base are manually analyzed and corrected, and a manual opinion processing flow and response strategy are formulated to form a new question-answer pair to be submitted to the system.

[0023] It's important to note that during the knowledge base retrieval phase, the system can set a similarity threshold (e.g., 0.8). If the highest similarity result retrieved is below this threshold, it's considered "not matched." At this point, the system automatically triggers a recording process, storing the full text of the current OA document, extracted key features, and the "no matching strategy found" status as a new record in a dedicated "pending issue database." The system backend can then notify the administrator. The administrator reviews the new record in the "pending issue database," performs manual analysis, and drafts the correct processing flow and response strategy. Then, the administrator submits this new knowledge (question-answer pair) through the knowledge base management tool. The system automatically calls the vectorization interface of the large language model to convert this new knowledge into a vector and incrementally updates it to the main knowledge base. In this way, when encountering a similar question again, the system can successfully match it, achieving self-learning and evolution.

[0024] Preferably, the knowledge base is built on the Tencent Intelligent Agent platform. In specific implementations, the "Tencent Intelligent Agent" platform service provided by Tencent Cloud can be directly used. The AI ​​intelligent agent question answering module interacts with the knowledge base and the large language model by calling the API of the Tencent Intelligent Agent platform.

[0025] Preferably, the generated OA response content includes the response approach, modification suggestions, draft amendments to the claims, and key points for debating the examination opinions.

[0026] It should be noted that during implementation, specific prompts can be designed to guide the large language model in generating content that meets the requirements. The instruction given to the large language model could be: "Based on the matched processing strategy and the current OA content, generate a detailed response to the OA, including the following parts: 1. Response approach (overall response strategy); 2. Suggested amendments to the application documents (e.g., specification, claims); 3. Draft specific amendments to the claims; 4. Points for rebuttal to each of the examiner's arguments." The large language model will then organize its language according to this instruction and generate a complete draft response to the OA covering all the above points. Example

[0027] See Figures 1-2 The aforementioned intellectual property application process management system supports the dynamic generation of patent application forms from multiple countries, specifically including the following steps: S301. Analyze the form content of patent application forms in different target countries, break down the form content into different reusable logical units, set up corresponding standardized controls for each logical unit and integrate them into the editor; S302. Using the editor, different logical unit controls are combined into patent application form templates that adapt to the requirements of different target countries, and corresponding template configuration data is generated based on the patent application form templates. S303. When the patent application form template of a specific target country is accessed, a layered information collection interface adapted to the input habits of Chinese users is generated according to the template configuration data corresponding to the patent application form template. The layered information collection interface contains two independent collection sub-units, namely the applicant and inventor information collection unit and the invention information collection unit, which are used to collect basic patent application data in Chinese format. S304. The basic data of Chinese format patent applications collected by the applicant and inventor information collection unit and the invention information collection unit are configured according to the template to automatically generate a patent application form that conforms to the official specifications of the corresponding target country.

[0028] In this embodiment, by splitting logical units and integrating them into controls, form elements of patent application forms from different countries can be reused, avoiding the repetitive work of developing separate forms for each country and improving template configuration efficiency. Furthermore, by configuring data through templates, flexible adaptation to form specifications of different countries can be achieved. When adjusting the form rules of the target country, no modification to the underlying system code is required; only configuration adjustments are needed for iteration. Simultaneously, by adapting the data collection interface to Chinese user habits, the barrier to entry for users directly filling out complex forms of the target country can be avoided, effectively reducing the error rate. In addition, by automatically converting data to generate application forms for the target country, no manual formatting or field reorganization is needed, effectively improving the efficiency of patent application form creation.

[0029] Furthermore, to facilitate user access, a patent application form template management interface can be configured. This interface provides access interfaces for patent application templates from multiple target countries. Users can initiate access to the patent application form templates for a specific target country simply by selecting the desired access interface.

[0030] In some examples, assuming the goal is to facilitate a Chinese company's patent application to the USPTO. First, the system pre-collects USPTO-issued patent application forms, functionally dividing the form content into two reusable logical units: inventor information and invention information. Standardized input controls are developed for each logical unit and integrated into a visual editor. Second, users can drag and drop these two logical unit controls in the visual editor to create a US patent application template. The system automatically generates corresponding template configuration data, which can pre-define the mapping and conversion rules between Chinese user input fields and USPTO-required fields. Third, when a Chinese company accesses the US patent application template, the system recreates a fully Chinese, layered information collection interface based on the configuration data. Users fill in basic data such as applicant name, address, application date, and inventor name according to Chinese conventions, without needing to understand USPTO form specifications. Finally, the system automatically reorganizes and converts the user-entered Chinese-formatted data according to the configuration rules, such as adjusting the address to the USPTO-required order and converting the date to US format, automatically generating a patent application form that conforms to USPTO specifications.

[0031] Preferably, the template configuration data pre-records mapping rules and data format conversion rules between Chinese input fields and target country form fields that are adapted to the requirements of the target country's form. When automatically generating a patent application form that conforms to the official specifications of the corresponding target country according to the template configuration data, the specific steps are as follows: the fields are reorganized and the format is converted according to the mapping rules and format conversion rules recorded in the template configuration data to automatically generate a patent application form that conforms to the official specifications of the corresponding target country.

[0032] Preferably, the step of splitting the form content of patent application forms from different target countries into different reusable logical units includes: hierarchically splitting according to the field attributes, layout logic, and required / optional attributes of the official form of the target country, dividing functionally related continuous fields into the same logical unit, with each logical unit corresponding to independent field validation rules, display logic, and data storage path, and synchronously recording the correspondence between each logical unit and the fields of the official form of the target country during the splitting process.

[0033] It's important to note that the logical unit splitting rules can resolve issues like field mismatches and omissions that easily occur during template assembly, ensuring the functional independence and field correspondence of each logical unit and improving the accuracy of template configuration. Continuing with the example of a Chinese company applying for a patent with the USPTO, during the splitting of the USPTO patent application form, the system hierarchically splits the form according to the target country's form field attributes, layout logic, and required / optional attributes. It divides functionally related consecutive fields such as "applicant name / title," "address," "contact email," "contact phone number," and "inventor name" into the same applicant and inventor information logical unit. This logical unit is configured with independent field validation rules (e.g., contact email must conform to email format, Chinese address must include province, city, and district information), Chinese display logic (all field prompts are in Chinese), and data storage path (all applicant and inventor information is stored in the applicant and inventor information database tables). The splitting process simultaneously records the correspondence between the logical unit and the USPTO official form fields. When assembling templates, the system automatically checks whether the applicant and inventor information logical units have been added; if missing, a prompt appears to prevent the generated template from lacking necessary fields.

[0034] Preferably, the template configuration data is stored in a distributed structure, including two independent storage partitions: a general rule base and a country-specific personalized rule base. The general rule base stores field mapping and format conversion rules common to all target countries, while the country-specific personalized rule base stores form adaptation rules unique to the corresponding target country. When the official form rules of the target country are adjusted, the content in the corresponding country-specific personalized rule base is updated to complete the template iteration.

[0035] It should be noted that the distributed storage structure of the template configuration data can solve the problem of needing to modify the entire template and incurring high iteration costs when adjusting the form rules of the target country. It improves the efficiency of template updates, and rule iteration can be completed without pausing system services, greatly reducing system maintenance costs.

[0036] The template configuration data can be stored in a cloud-distributed system, divided into two independent storage partitions: a general rule base and a country-specific personalized rule base. The general rule base stores rules for "special character escaping, required field validation, and sensitive word filtering" that are common to all target countries. The country-specific personalized rule base stores form adaptation rules unique to different target countries, such as the United States, the European Union, and the United Kingdom. When a target country modifies its option rules, the operations and maintenance personnel only need to update the corresponding content in the country-specific personalized rule base, and the patent application template for the target country will take effect immediately without modifying the editor's underlying logic or the content of the general rule base, or pausing the system service.

[0037] Preferably, in the applicant and inventor information collection unit and the invention information collection unit, each collection unit supports step-by-step filling, real-time verification, and automatic temporary storage functions. During the collection process, Chinese prompts and Chinese general format examples are used to guide the user's input.

[0038] It should be noted that the hierarchical sub-unit structure of the data collection interface can solve the problems of disorganized information and easy omissions when users fill in the information, improve the accuracy of information collection, shorten the filling time, and lower the threshold for users to fill in the information. Among them, the hierarchical information collection interface adapted for Chinese users includes two independent collection sub-units: the applicant and inventor information collection unit prompts users to fill in the applicant's name, Chinese address, contact information, inventor's name, etc. in Chinese according to Chinese habits; the invention information collection unit prompts users to select the patent category to be applied for; each collection unit supports step-by-step filling, real-time verification, and automatic saving functions. After exiting, users can continue from the last entry position when they re-enter. All prompts are in Chinese, so users do not need to understand the form requirements of the target country.

[0039] Preferably, when N co-inventors need to be added, the applicant and inventor information collection unit generates an independent inventor information collection subpage for each co-inventor. After collection, the information of multiple inventors is sorted, merged, and formatted according to the requirements of the target country's form, and automatically filled into the corresponding co-inventor field position in the target country's form.

[0040] It's worth noting that dynamic adaptation for multiple co-inventors solves the problem of manually organizing, sorting, and formatting inventor information in multi-inventor scenarios, improving processing efficiency and avoiding application risks caused by manual sorting errors. For example, when a Chinese company and two affiliated companies jointly apply for a US patent, the user clicks the inventor editing icon on the data collection interface. The system automatically generates a separate subpage for collecting inventor information, where the user fills in the inventor's Chinese address and other information. After submission, the system automatically sorts the information in the order of first inventor, co-inventor 1, and co-inventor 2, converting all Chinese-format inventor information to US format and automatically filling it into the corresponding co-inventor field in the USPTO form, without requiring the user to manually adjust the information order or format.

[0041] Preferably, the process of reorganizing and converting fields according to the mapping rules and format conversion rules recorded in the template configuration data includes: identifying the field types and content attributes of the collected basic data of Chinese format patent applications, matching the corresponding mapping rules in the template configuration data, and directly completing the field content migration if there is a one-to-one mapping relationship, and completing the content reorganization according to the preset field merging and splitting rules if there is a many-to-one or one-to-many mapping relationship.

[0042] It should be noted that matching field mapping and format conversion can solve the problem of content mismatch that easily occurs when fields are mapped in a many-to-one or one-to-many manner, improving the accuracy of data conversion and reducing the application risk caused by conversion errors. For example, when the system recognizes the two independent fields "China" entered by the user, it matches them with the many-to-one mapping rules of the target country's format fields in the template configuration data, and automatically merges the contents of the two fields into the target country's format fields and fills them into the corresponding positions.

[0043] Preferably, after the step of automatically generating a patent application form that conforms to the official specifications of the target country, a dual verification step is also included. The first verification is a format verification, which automatically checks whether the fields, format, and layout of the generated form fully conform to the official requirements of the target country. The second verification is a content verification, which automatically compares whether the generated form content is consistent with the original content entered by the user. After all verifications pass, the form can be downloaded.

[0044] It should be noted that the dual verification after the application form is generated can resolve application risks caused by conversion errors and content omissions, thereby improving the compliance rate of the application form. For example, after the system automatically generates the US patent application form, it first initiates the first level of format verification, automatically checking whether the number of fields, layout, date format, and address format of the generated form comply with the latest official requirements of the USPTO. If there are format errors, they are automatically corrected and a correction log is generated. Then, the second level of content verification is initiated, automatically comparing whether the converted inventor's name, patent name, etc., are completely consistent with the content originally entered by the user. If there are content differences, a difference prompt will pop up to guide the user to confirm. Only after both verifications pass can the user be allowed to download the application form.

[0045] It should be noted that the above embodiments are merely preferred embodiments of the present invention, and the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention, and the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An OA response method based on an AI agent in an intellectual property application process management system, characterized in that, include: In the intellectual property application process management system, an AI intelligent agent question-answering module and a knowledge base for storing OA document process questions and answers are created. The AI ​​intelligent agent question-answering module is configured so that it can call a large language model to access the knowledge base. The AI ​​intelligent agent question-answering module obtains the OA document and calls the large language model to automatically identify the OA document, extract preset key features, and retrieve the opinion processing flow and response strategy related to the key features from the knowledge base; Based on the matched opinion processing flow and response strategy, the system automatically generates response OA content through the large language model, and provides the response OA content to the internal administrator of the intellectual property application process management system through the AI ​​intelligent agent question-and-answer module.

2. The OA response method based on the AI ​​agent of the intellectual property application process management system according to claim 1, characterized in that, Create a knowledge base for storing OA document process Q&A, including: Collect the question and answer content of OA documents in the field of intellectual property, extract question and answer features, and form structured question and answer pairs; The text content of the question-and-answer pairs is imported into the knowledge base. The large language model is called, and the vector model is used to transform the text content of the question-and-answer pairs imported into the knowledge base. The text of the question-and-answer pairs is converted into vector representations and stored in the knowledge base to obtain a knowledge base for storing OA document process questions and answers.

3. The OA response method based on the AI ​​agent of the intellectual property application process management system according to claim 1, characterized in that, Obtaining OA documents through the AI ​​intelligent agent question-and-answer module includes: automatically communicating with the official server of intellectual property application management according to preset rules through the communication interface of the AI ​​intelligent agent question-and-answer module, and obtaining new OA documents periodically or in real time.

4. The OA response method based on the AI ​​agent of the intellectual property application process management system according to claim 3, characterized in that, The user interface of the AI ​​agent question-and-answer module is a mini-program interface, APP interface, or web page interface that provides agent chat functionality.

5. The OA response method based on the AI ​​agent of the intellectual property application process management system according to claim 1, characterized in that, The large language model automatically recognizes and understands OA text content in multiple languages, including Chinese.

6. The OA response method based on the AI ​​agent of the intellectual property application process management system according to claim 1, characterized in that, The key features include application number, type of examination opinion, legal clauses, comparison of technical features, obviousness assessment, and routine technical assessment.

7. The OA response method based on the AI ​​agent of the intellectual property application process management system according to claim 6, characterized in that, The AI-powered question-and-answer module presents the OA (Office Automation) responses to the internal administrator in a structured format.

8. The OA response method based on the AI ​​agent of the intellectual property application process management system according to claim 6, characterized in that, If no matching opinion processing flow and response strategy related to the key features is found in the knowledge base, the content of the OA document and the question obtained this time will be automatically recorded into the knowledge base. For issues recorded in the database, manual analysis and correction are conducted, and a process for handling manual opinions and a response strategy are developed to generate new question-answer pairs for submission to the system.

9. The OA response method based on the AI ​​agent of the intellectual property application process management system according to claim 6, characterized in that, The generated OA response includes the response approach, suggested amendments, a draft amendment to the claims, and key points for rebuttal to the examination opinions.

10. The OA response method based on an AI agent in an intellectual property application process management system according to any one of claims 1-9, characterized in that, The intellectual property application process management system supports the dynamic generation of patent application forms from multiple countries, specifically including the following steps: Analyze the form content of patent application forms in different target countries, break down the form content into different reusable logical units, set corresponding standardized controls for each logical unit and integrate them into the editor; The editor combines different logical unit controls into patent application form templates that adapt to the requirements of different target countries, and generates corresponding template configuration data based on the patent application form templates. When the patent application form template of a specific target country is accessed, a layered information collection interface adapted to the input habits of Chinese users is generated according to the template configuration data corresponding to the patent application form template. The layered information collection interface contains two independent collection sub-units, namely the applicant and inventor information collection unit and the invention information collection unit, which are used to collect basic patent application data in Chinese format. The basic data of Chinese format patent applications collected by the applicant and inventor information collection unit and the invention information collection unit are configured according to the template to automatically generate a patent application form that conforms to the official specifications of the corresponding target country.

Images