21 results about "Technical documentation" patented technology

The invention discloses a method for analyzing a matching degree between a demand and an output result based on text semantics. The method comprises the following steps: step 1, labeling a data set; step 2, technical document preprocessing; 3, training and predicting a single-parameter model; 4, integrating prediction results of the multi-parameter model; the method has the beneficial effects thatthe method is simple; deep learning and the NLP technology are applied to the field of project association degree calculation of enterprise project management for the first time. Calculating an association matching degree between the two projects according to project requirements and result description; the associated project positioning difficulty is effectively reduced; meanwhile, the demand side can be helped to quickly and efficiently locate high-quality projects adapting to the demand of the demand side; time and resource investment for achievement screening and matching are greatly reduced, the association matching degree between projects is calculated by means of text data of existing project achievement technical documents and project declaration guidelines, and then large enterprises are assisted in screening high-quality projects with the high matching degree in the project bidding and tendering link.

The application relates to the technical field of distributed applications, and discloses a document synchronization method for code change, which comprises the following steps: in response to a code deployment request, monitoring code change through a change capturing engine and obtaining a unified change set; performing document paragraph positioning on an interface definition file according to the unified change set, constructing a change-paragraph mapping list and a document synchronization task; performing acceptance review on the document synchronization task based on acceptance rules to obtain a review result; and making a decision based on a blocking condition and the unified change set, the change-paragraph mapping list and the review result to determine whether to block code deployment and complete document synchronization. The method can realize accurate positioning of abstract code change to technical document paragraphs, form a unified and feasible evidence chain to jointly drive a document synchronization process and decision-making. The application further discloses a document synchronization device for code change and electronic equipment.

The application provides a technical document automatic generation method, system, device and storage medium, the method comprises the following steps: analyzing a demand document through a feishu intelligent slice, and extracting complete content; an intelligent agent actively inquires a user to obtain an interface name, an application ID and an interactive picture link, triggers corresponding MCP to obtain interface data and interactive documents, and performs integrity check; in combination with demand, interface and interactive content, the intelligent agent inquires technical stack information, and generates a standardized technical document by using a prompt word engineering. Then, a double check mechanism is started: the intelligent agent can automatically correct the error, and the user is asked to confirm the ambiguous demand until the document is correct; finally, the feishu document generation MCP is triggered, the demand, interface, interaction and final document are integrated into a Markdown format, and a link is pushed to an AI programming tool for generating code. The application can realize full-process automatic conversion from a demand document to a technical document in a Web webpage development scene, and provide a standardized technical basis for an AI programming tool.

This application relates to a document generation method, apparatus, device, storage medium, and program product. The method includes: constructing an outline generation prompt text based on a first tree structure composed of the file paths of various technical documents of a target project; generating an outline based on the outline generation prompt text using a large language model to obtain the outline of the technical document for the target project and the chapter description information of each initial chapter in the outline; recursively splitting each initial chapter according to the chapter description information to obtain the chapter content of each sub-chapter under each initial chapter; and generating the technical document for the target project based on the outline and the chapter content of each sub-chapter under each initial chapter. This method can improve the accuracy of generating technical documents for the target project.

This specification provides a technical document question-and-answer pair generation, evaluation, and optimization system and method. The system includes: a document reading module for reading structured technical documents; a multi-level heading parsing module for dividing the technical document into several content blocks according to the document heading hierarchy rules; a question-and-answer generation module for generating questions and answers (question-and-answer pairs) based on the content blocks using natural language processing technology and a large language model; a question-and-answer evaluation module for performing quality scoring and consistency checks on the question-and-answer pairs and outputting evaluation results; a question-and-answer optimization module for optimizing and correcting the question-and-answer pairs based on the evaluation results; and a batch processing and output module for outputting the optimized and corrected question-and-answer pairs using a multi-document batch processing method.

The disclosure relates to a system and method for unsupervised validation of Large Language Model (LLM)-generated outputs. The method includes the steps of receiving input data, which may comprise structured or unstructured text; generating an LLM output based on the input data; extracting a first set of topics from the input data, with each topic represented by a set of keywords; converting the first set of topics into human-readable reference data; extracting a second set of topics from the LLM output; and converting the second set of topics into human-readable candidate data. The method further includes comparing the reference data with the candidate data using one or more performance metrics. Finally, the method determines a validation score based on the comparison results. The disclosed method allows for the objective validation of LLM-generated content, enhancing accuracy and efficiency in diverse applications such as customer service, technical documentation, and legal analysis.

This invention provides a static defect analysis method and system for an agent development framework based on a large language model. The method includes: acquiring the agent development framework to be analyzed and the corresponding downstream applications; extracting the mapping relationship between modules and functions in the agent development framework from its technical documentation and constructing a module-function feature table; acquiring defect-related text describing the defects of the agent development framework in natural language, extracting initial semantic features reflecting the text's thematic structure, and inputting the defect-related text and the initial semantic features into a large language model after vectorization encoding to extract structured defect problems; constructing an abstract syntax tree for each downstream application corresponding to the agent development framework, identifying the actual function calls in each downstream application through the abstract syntax tree, locating the functions and modules associated with the actual defects, and generating static analysis results.

A system for managing a technical document for a storage device includes: a large language model (LLM) configured to receive parts selected from a first version of the technical document, which have a target characteristic, and a second version of the technical document newer than the first version of the technical document, and generate first vectors for the selected parts from the first version of the technical document and second vectors for all parts from the second version of the technical document based on semantics. A comparator compares each of the second vectors with the first vectors to find a closest first vector, and generates a similarity metric between each of the second vectors and the closest first vector. A classifier classifies each of the parts from the second version of the technical document based on the similarity metric and text values of the parts from the second version.

This invention provides a domain-adaptive pre-training and fine-tuning method and system for the gas turbine field. The method includes: constructing a multi-source fusion corpus for the gas turbine field, including publicly available standard data, internal technical documents, operation logs and operating condition data, and expert experience knowledge; based on a general large language model, using the corpus for hardware-accelerated domain-adaptive continuous pre-training on a heterogeneous computing platform to obtain a basic model with enhanced domain knowledge. During continuous pre-training, operating condition data is encoded into high-dimensional vectors, fused with the embedding vectors of text tokens, and then input into the model; using a gas turbine field instruction fine-tuning dataset, the basic model is subjected to supervised fine-tuning to obtain a dedicated large language model for the gas turbine field. This addresses the knowledge barriers and performance bottlenecks in industrial applications caused by the high hardware cost of general models and the lack of optimization and adaptation for domestic computing platforms.

Embodiments of the present disclosure provide methods, apparatus, systems, computing devices, and computing entities for displaying content found in technical documentation for an item via an IETM viewer. In accordance with one embodiment, a method is provided comprising: providing a window for display via the IETM viewer, the window displaying the content found in the technical documentation; receiving a focus of a first portion of the content received based at least in part on the user performing a first action; and in response, causing the first portion to be conveyed using a first enhancing format; and thereafter: receiving a focus of a second portion of the content received based at least in part on the user performing a second action; and in response: causing the first portion to no longer be conveyed using the first enhancing format; and causing the second portion to be conveyed using a second enhancing format.

The application provides a kind of AI knowledge base document construction method and system, method includes: scanning front-end project source code directory to identify technical document;Convert the first technical document of non-markdown identified into Markdown format;Extract the reference path of media resource in the second technical document of Markdown format, and establish the first mapping relationship between the absolute path of the second technical document and the relative path of the media resource, and the second mapping relationship between the absolute path of the media resource and the hash file name;According to the first and second mapping relationship, replace the media resource reference address that cannot be directly accessed in the second technical document;Add uniform prefix identifier to the title of different levels in the second technical document;The second technical document is packaged into network request and sent to the specified AI large model server, so that the AI large model receives the second technical document and stores it into AI knowledge base according to structure fragmentation.

The application discloses a retrieval method and device of application function information. It is related to the field of artificial intelligence, and the method comprises the following steps: receiving a retrieval request initiated by a requester; in response to the retrieval request, generating a reference specification set of a target application by using a first generation model, wherein the first generation model is obtained by training a first initial generation model by using an application attribute library and a technical specification library which are marked with application specification set samples of each application in a plurality of applications; generating a target specification set according to the reference specification set and a target technical document of the target application by using a second generation model, wherein the second generation model is obtained by training a second initial generation model by using application specification set samples which are marked with function specification set samples of each function in each application and function technical documents; and sending the target specification set to the requester. Through the application, the problem of low retrieval efficiency of application function-related technical specifications in the related art is solved.

We provide the system. [Solution] A means of receiving information input from a user and analyzing the key points of the information using natural language processing, A means for automatically collecting relevant technical documents from external data sources and internal record bases based on the analyzed key points, A means of constructing a virtual simulation under specific environmental conditions and performing related work verification, A method for generating proposed answers using artificial intelligence based on collected materials and the results of work verification, In order to provide the generated draft answers, there is a means by which experts can check the content and make final changes, A means of quickly providing users with the optimal solution in real time through an application installed on a smart device, A system that includes this.

This invention provides a data integration method for technical documents. The method includes: importing structured data into a content management system, wherein the structured data originates from at least two different data sources, and the content management system is configured to manage technical documents; associating specific data entries in the imported structured data with multiple dynamic placeholders in one or more technical documents managed by the content management system; populating the multiple dynamic placeholders in the one or more technical documents with corresponding values ​​from the imported structured data; and automatically updating the corresponding values ​​of the populated multiple dynamic placeholders in the one or more technical documents in response to an update of the structured data. This method can significantly improve the efficiency and accuracy of technical document writing and greatly reduce maintenance costs by automating the import, population, and updating of dynamic placeholders.

The application provides a test point generation method and device, electronic equipment and storage medium, relates to the chip design and manufacturing technical field, and includes the following steps: reading a technical document related to a chip to be tested from a memory; identifying technical features in the technical document, and classifying the technical features to corresponding technical dimensions based on a preset feature classification rule; calling test point generation rules corresponding to each technical dimension, processing the technical features under each technical dimension, and generating test point data under each technical dimension; and aggregating the test point data under each technical dimension and outputting a verification test point file of the chip to be tested. The method and device provided by the application ensure that the analysis of the chip design can cover all aspects through the decomposition of multiple technical dimensions, realize comprehensive generation of test point data, and greatly improve the completeness and efficiency of pre-simulation verification of the chip.

There is provided an information processing device that includes a generation unit configured to generate an instruction text for instructing a generative model and a presentation unit configured to present an output from the generative model. The generation unit is configured to generate a first instruction text for instructing a generative model to extract a portion where a technical feature is unidentifiable in a text related to a scope of a patent right, and generate a second instruction text for instructing the generative model to extract a correspondence relationship with a technical document relevant to the text related to the scope of the patent right with respect to the portion extracted by the generative model, based on the first instruction text. The presentation unit is configured to present the correspondence relationship extracted by the generative model based on the second instruction text.

This invention discloses an intelligent PDF document processing system and method based on a multimodal large language model, belonging to the field of artificial intelligence and document processing technology. The method includes: receiving a PDF document and uploading it to a multimodal large language model service platform to obtain a file identifier; extracting text content through multi-turn dialogue; extracting embedded images using a multi-level degradation strategy; scanning table fragments, performing sequence number continuity detection on adjacent tables, locating technical parameter regions, extracting parameter content, and performing text cleaning to remove redundant information; and saving the cleaned parameters to a designated location. This invention achieves fully automatic multimodal parsing and structured extraction of PDF technical documents, with high image extraction coverage, automatic merging of cross-page tables, reduced noise in parameter extraction for mixed Chinese and English text, and no manual intervention required throughout the entire process. It can be widely applied to technical document processing in industrial fields such as new energy batteries and electronic components.

This application provides a method, apparatus, device, and medium for extracting business knowledge based on an AI model. It extracts information from business data in a business system to generate initial business knowledge units. The business data includes source code, related technical documents, and auxiliary materials. A multi-dimensional credibility assessment is performed on the initial business knowledge units to determine their initial credibility level. Initial knowledge units whose initial credibility levels do not meet preset requirements are repaired to obtain target business knowledge units and their corresponding target credibility levels. A business knowledge base is constructed based on the target business knowledge units and their corresponding target credibility levels. Using an AI model as the core, and through techniques such as joint parsing of code and documents, stability verification of minor code perturbations, and comprehensive verification of business logic, a systematic guarantee of the accuracy, completeness, and consistency of the business knowledge extraction results is achieved.

Disclosed herein are systems and method for machine learning (ML)-assisted analysis of scientific or technical documents and associated software code, the method comprising: retrieving a scientific or technical document and a software code associated with the scientific or technical document, the software code including executable code and / or source code; analyzing the scientific or technical document using a trained paper analysis ML model configured to identify subject matter, logic, algorithms, and / or parameters of the scientific or technical document; analyzing the software code using a trained software analysis ML model configured to identify subject matter, logic, algorithms, and / or parameters of the software code; and comparing the algorithms and parameters of the scientific or technical document with the algorithms and parameters of the associated software code to identifying corresponding algorithms and parameters in the scientific or technical document and the software code.

This application discloses a method, medium, and product for searching technical documents. The method includes: acquiring the original technical document; performing structured parsing on the original technical document to identify multiple semantic blocks and their hierarchical relationships; constructing a document structure tree based on the multiple semantic blocks and hierarchical relationships, wherein each node in the document structure tree corresponds to a semantic block, and each non-root node is associated with path information of all its ancestor nodes; constructing a multi-level composite index based on the nodes in the document structure tree, including a first type index and a second type index; responding to a received query statement, performing searches in the first and second type indexes respectively to obtain multiple candidate result nodes; and reordering the multiple candidate result nodes based on their positions in the document structure tree to generate search results.