103 results about "Semantic relation" patented technology

This invention discloses a 3D scene segmentation and interaction method based on 3D Gaussian, comprising: Step 1, instance discovery: inputting 3D Gaussian scene data, dividing the 3D Gaussian points into structurally coherent instance-level Gaussian groups to obtain refined instances; Step 2, instance and scene semantic assignment: selecting representative viewpoint images of refined instances and inputting them into a visual-language model to generate semantic description labels for the instances; filtering instance pairs in the scene through spatial geometric relationships, and clarifying the spatial relationship description of instance pairs through a large model, constructing a static scene graph that integrates geometric proximity relationships and instance semantic relationships, forming a structured description of all instances; Step 3, natural language-driven instance localization and interaction: receiving user input commands, combining the instance semantic description, the static scene graph, and the real-time viewpoint direction relationship during the query to perform multi-dimensional matching, determining the target instance ID, and executing interactive operations.

This invention discloses a method and system for extracting and analyzing key points of legal documents based on a large model, relating to the field of natural language processing technology. The invention generates a semantic parsing candidate graph based on legal semantic nodes and semantic relationships, calculates the matching scores of candidate paths in the semantic parsing candidate graph, summarizes the matching scores of all candidate paths in the semantic parsing candidate graph, and calculates the topological consistency index of the semantic parsing candidate graph. It identifies semantic conflict units and calculates initial conflict weights, and calculates the overall conflict impact of the semantic parsing candidate graph. Based on the topological consistency index, the basic semantic confidence of the semantic parsing candidate graph, and the overall conflict impact, it calculates a comprehensive score for the semantic parsing candidate graph and selects the optimal result, comprehensively optimizing and dynamically adjusting the legal semantic parsing results, significantly improving the accuracy, reliability, and intelligence level of legal text parsing.

Technologies for machine learning-based entity understanding and resolution are disclosed. Data for training an entity resolution model is collected to learn semantic relationships associated with entity names. The entity names are provided in a domain of documents that follow the semantic conventions differently from natural language semantic conventions. The data includes entries each specifying an entity name and a label. The entity resolution model is trained using the data to learn and generalizes the semantic relationships and is deployed to serve requests for resolving an entity name from text extract from a document image.

The application relates to the technical field of intelligent reasoning and safety judgment, and discloses a high-speed rail platform safety judgment method and system based on mixed-precision reasoning, which comprises the following steps: acquiring multi-source semantic observation records and generating a safety observation element set; constructing a platform safety relation graph; determining relation conflict density, closed residual error, cross-source divergence degree and reasoning difficulty level; generating a mixed-precision bit width scheduling table; performing graph relation reasoning to obtain an initial safety judgment vector and an initial judgment boundary quantity; in step 6, a final judgment vector is determined; and in step 7, a locked safety level is determined and a safety judgment package is output. The application realizes mixed-precision safety judgment and locked output driven by multi-source semantic relation of a high-speed rail platform.

The application discloses a knowledge graph construction method and device of a promotion path, wherein the method comprises the following steps: step S1, combing core concepts related to the promotion path and semantic relations thereof, constructing a basic knowledge framework, and creating an ontology model based on the basic knowledge framework; step S2, setting mapping rules between ontology data and a graph database, and mapping the ontology model to the graph database; step S3, collecting multi-source information, performing knowledge extraction on the multi-source information to obtain knowledge triples and constructing entities, and constructing a knowledge graph by comprehensively combining the ontology model and the entities; and step S4, performing retrieval and correlation analysis based on the knowledge graph, forming promotion path instances, dynamically updating the promotion path instances to the knowledge graph, and obtaining a promotion path knowledge graph. The promotion path knowledge graph is obtained based on dynamic updating of the knowledge graph.

This invention relates to the field of software development technology and discloses a document processing method, apparatus, device, and medium. The method includes: acquiring multiple heterogeneous data sources of a software project, wherein the heterogeneous data sources include at least source code and software documents associated with the source code; constructing a semantic difference graph describing the semantic relationships between the heterogeneous data sources based on the multiple heterogeneous data sources; assessing the synchronization risk value between the software documents and the current source code based on the semantic difference graph; if the synchronization risk value exceeds a preset threshold, determining the target content in the software documents that needs to be updated synchronously based on the semantic difference graph; generating differential update information for updating the target content based on the current source code and the semantic difference graph; and triggering an update processing flow for the software documents based on the differential update information. This method can be applied to document processing scenarios in fintech and healthcare, improving the accuracy and efficiency of document processing.

The application belongs to the field of natural language processing, and particularly relates to a method and system for paraphrase recognition based on semantic primitive knowledge and abstract semantic representation, which comprises the following steps: performing word segmentation on a sentence, and performing word-level vector representation and semantic primitive knowledge representation; performing mean value processing on the semantic primitive knowledge representation result, and extracting interactive attention feature information of the mean value processing result by using global semantic information to obtain global semantic primitive representation; performing abstract semantic analysis on a to-be-recognized paraphrase sentence from a sentence structure to obtain a single-root directed acyclic graph, and performing global semantic primitive representation and word-level vector representation; extracting global and local feature information in the order of the directed acyclic graph, and performing distance feature measurement on the information; inputting the distance feature measurement result into a neural network to obtain a recognition result; the application introduces external semantic primitive knowledge to perform semantic representation, the accuracy of the semantic primitive knowledge representation is assisted by global semantic information, and the abstract semantics of a Chinese paraphrase sentence is analyzed to obtain semantic relations, so that the accuracy of paraphrase recognition is improved.

The present disclosure sets forth systems, apparatuses, and methods that uniquely combine vector-based semantic processing with graph-based knowledge representations, leveraging the strengths of both approaches to improve similarity searches and context matching and capture nuanced semantic relationships between concepts. The disclosed systems, apparatuses, and methods include components configured to receive data, determine based on the data one or more entities, and generate, based on the one or more entities, a probability distribution, context associated with the data, and historical interactions, one or more semantic mappings comprising nodes interconnected according to weighted relationships, and update, based on at least one of additional data, context, or interactions, the one or more semantic mappings.

This invention discloses a method and system for on-site monitoring and early warning based on retrieval enhancement. The method involves acquiring multimodal data from on-site inspections, transforming this data into several independent semantic segments based on its acquisition attributes, preprocessing these segments to obtain on-site state representation vectors, filtering and identifying valid document fields in a document library, performing natural language vector transformation on the valid document fields to obtain knowledge segment vectors, constructing a trusted knowledge segment vector library based on the document library's updates, retrieving matching knowledge segment vectors from the trusted knowledge segment vector library based on the on-site state representation vectors, and performing early warning operations for on-site anomalies based on the semantic relationships between the matching knowledge segment vectors. By forming semantic representation vectors from on-site inspection multimodal data and knowledge segment vectors from a benchmark document library, the method achieves matching retrieval between vectors, resulting in standardized early warning operations.

This invention provides a semantic segmentation method and system for human skin and clothing. The method includes: constructing a pixel-level mutually exclusive semantic category system containing multiple refined human body part categories and clothing categories, and labeling a human image dataset accordingly; constructing a multi-category output multi-part semantic segmentation model, and training it using a joint loss function; the joint loss function includes not only a pixel-level cross-entropy loss term to drive the model to accurately distinguish each fine-grained category, but also a mutually exclusive regularization term, which imposes mathematical constraints to force the model to learn and adhere to the semantic relationship between the two major categories of skin and clothing. This invention, through the above method, can simultaneously achieve high-precision multi-part skin segmentation and clear skin-clothing boundary delineation, solving the problems of coarse segmentation granularity, blurred boundaries, and semantic confusion, providing high-fidelity segmentation results for downstream applications.

This invention relates to the field of semantic understanding technology, specifically to a method for semantic understanding and intelligent question answering in online courses for mobile terminals. The method includes the following steps: parsing the question text to identify subject-verb-object semantic relationships and generate corresponding structural expressions; collecting concept and behavioral terms from a knowledge base for semantic comparison and generation of course semantic expressions; reading positional analysis to generate question relationship expressions based on concept associations; collecting subtitles and knowledge structures for comparative association and generation of knowledge correspondence expressions; and extracting subtitle fragments and integrating playback positions to generate a processing scheme. In this invention, deep semantic comparison and positional analysis are performed by linking object terms in the course knowledge base, overcoming the limitations of shallow character matching to achieve deep question intent restoration. Simultaneously, subtitle content and chapter structure are collected for contextual analysis, and subtitle fragments are deeply integrated with video playback progress to construct a response mechanism that highly fits the actual teaching context, completely changing the isolated retrieval mode and eliminating interaction breaks.

This disclosure provides a video processing method and related equipment. The method includes acquiring a video to be processed, dividing the video to be processed into a plurality of scene video segments based on scene transition positions, determining a text description corresponding to each of the scene video segments, and removing the scene video segments having overlapping semantics in the video to be processed based on the semantic relationships between the text descriptions to obtain a target video.

This invention discloses a bolt defect detection method that integrates component semantic relationships and multi-task learning, relating to the fields of power line inspection and machine vision inspection. This invention defines bolt defect types and transforms them into geometric semantic attributes, establishing mapping rules between defect types and geometric semantic attributes, topological rules for bolt semantic relationship graphs, and semantic consistency constraint rules. A multi-task deep learning network is constructed and trained using a total loss function to extract multi-scale visual features of bolts. A semantic relationship graph is constructed using bolt components as graph nodes. Relationship weights of graph nodes are calculated through a relationship-aware network. Enhanced node features are obtained after message passing and feature fusion through a graph attention mechanism. Bolt defect classification, semantic attribute regression, and component segmentation tasks are performed in parallel, ultimately outputting the bolt defect detection results. This invention significantly improves the accuracy and interpretability of bolt defect detection, providing a reliable technical solution for intelligent inspection of bolts in power transmission lines.

The present application belongs to the technical field of data processing, and particularly relates to a flowchart understanding method, system, device and storage medium. The present application aims to solve the technical problems of the prior art, such as dependence on a large amount of high-quality artificial annotation data and lack of accurate modeling of high-order semantic relationships. The present application comprises the following steps: data analysis and data cleaning are performed on original flowchart data to obtain flowchart text data; a training data set is constructed, and a pre-trained visual multi-modal large model is fine-tuned through the training data set; flowchart entities and their logical relationships are extracted from the flowchart text data by using a flowchart domain large model, a structured flowchart knowledge graph is constructed, and the flowchart knowledge graph is stored in a graph database; a user question is obtained, relevant entities and relationships are retrieved in the graph database, and a preliminary retrieval answer is obtained; the preliminary retrieval answer and the user question are fused to obtain a prompt question; and the prompt question is input into the flowchart domain large model to obtain a final flowchart understanding answer.

The application provides a text-to-SQL training data generation method based on knowledge analysis driving, and belongs to the fields of artificial intelligence and database technologies.The method proposes: constructing an explicit structured knowledge base for storing domain rules, effective operations and semantic relationships extracted from a target database schema;under the guidance of the knowledge base, generating natural language question-SQL pairs and synchronously outputting explicit reasoning chains recording the basis for decision making; performing semantic verification by comparing the explicit reasoning chains with semantic constraints in the knowledge base to identify potential semantic violations regardless of whether SQL execution is successful or not; if a semantic violation is detected, accurately locating the error step according to the reasoning chain and making targeted correction according to the knowledge base.The application realizes the transparency and verifiability of the generation process by introducing explicit reasoning chains, and significantly improves the semantic consistency and quality of the synthesized data.

A human problem-solving thinking driven embedded enhancement method for logical reasoning problems, the present application relates to a multiple-choice logical reasoning question and answer problem in the field of natural language processing, which aims to solve the technical problems of unclear text structure analysis, insufficient use of problem related elements and difficult logical reasoning problem semantic mining in the existing logical reasoning question and answer problem solving method. The method mainly includes data preprocessing, arranging the sample into a text sequence suitable for neural network model processing, designing a question-based embedding update module based on human problem-solving thinking, first initializing character embedding representation, using a pre-trained language model to vectorize the characters in the sample, then using the embedding sequence of the question part to construct a logical information filtering vector, and then updating the embedding representation of the stem and options using the filtering vector, and a semantic relationship modeling module referring to human reading habits, analyzing the basic semantic unit of the neural network.

This application discloses a PLM persistent storage and semantic reasoning method for CAD / CAE / CAM collaboration. The method includes: obtaining a submission request; hashing the content to be written in the submission request into blocks and writing it to content storage according to the content hash value; generating an expression record and recording it in a cross-domain expression set, pointing to the old expression record through a derived pointer; generating semantic annotation records and explicit semantic relationship records based on the target object identifier contained in the submission request; performing consistency verification; after passing the verification, performing incremental ontology reasoning and rule reasoning based on the semantic subgraph range limited by the change set, generating inferred semantic relationships and interpretation chains; generating a submission record and appending it to the metadata storage and updating the index; performing impact analysis based on the expression dependency graph, generating recalculation or repair markers, publishing change events, and triggering collaborative response actions. The advantage of this application is that it significantly reduces the cost of cross-domain collaboration interruption chain rework, storage redundancy, and manual intervention.

This invention discloses a method for constructing a data semantic model based on a large AI model. The method involves semantic extraction, semantic object formation, semantic consistency judgment and semantic constraints, semantic relation generation, and semantic model verification and evolution using a continuous algorithm chain based on a large AI model. The method is characterized by the following steps: (1) semantic input and semantic fragment extraction; (2) semantic object candidate generation; (3) semantic consistency judgment and object shaping; (4) semantic object comparability judgment and constraint generation; (5) semantic relation inference and relation shaping; and (6) semantic model generation and closure verification.

The application provides a multi-dimensional cognitive state vector simulation agent digital textbook dynamic adaptability evaluation method, and relates to the technical field of artificial intelligence assisted education. The method comprises the following steps: extracting teaching entities and semantic relationships between the entities in a source file of a digital textbook to be evaluated through a multi-modal analysis model, and constructing a knowledge graph; initializing multiple groups of virtual learning agents based on a pre-trained large language model, and configuring differentiated cognitive state vectors for each group of virtual learning agents; driving each group of virtual learning agents to traverse the teaching knowledge graph according to a topological sorting by using a knowledge boundary mask mechanism, obtaining understanding feedback of each node by each group of virtual learning agents, and generating confusion degrees and thought chains in the understanding feedback process; calculating a cognitive block index and an adaptability distribution curve of the source file of the digital textbook by using a weighted summation algorithm, and generating an evaluation report. The application simulates human cognitive processes by using a large language model agent, and evaluates the quality and adaptability of a digital textbook.

The application discloses a natural resource data association and graph display method and device and a storage medium, and relates to the technical field of geographical entity space identity coding. The method comprises the following steps: generating a space identity code of a geographical entity, wherein the space identity code is used for uniquely identifying the geographical entity in a global range; based on a preset association rule, an association relationship between the space identity code and original codes of at least one natural resource business data is established, and association data is obtained; according to a natural resource business process, a semantic relationship between the association data is defined; and based on the association data and the semantic relationship, a knowledge graph reflecting the space-time change of natural resources is constructed and displayed. The scheme combines the characteristics of geographical entity space identity coding, such as space position determination, global uniqueness and space-time association, and connects multi-source space-time data across businesses, systems and departments through geographical entity space identity coding, so that all-chain data falls on the same space unit.

The application provides a text classification method and system based on guided graph contrastive learning. The text classification method comprises: performing upper and lower semantic modeling on an input original text, converting a discrete text sequence into a continuous vector representation; mining potential theme semantic structures in the text based on a neural network to obtain distribution characteristics of the document in a theme space; representing the text data in a graph structure form to depict multi-level semantic relationships and structural relationships among the document, words and themes; generating multiple enhanced text graph views with consistent semantics but different structures through guided disturbance on the graph structure; and training in a graph contrastive learning mode to maximize the consistency of the representation of the same document under different enhanced text graph views and to reduce the distance between the representations of different documents. The method significantly improves the generalization performance and robustness of the model in the text classification task and reduces the dependence on artificial feature design and large-scale labeled data.

The application discloses a kind of publishing field knowledge service construction methods based on knowledge graph, it is related to intelligent digital publishing technical field, including, according to publishing content structure node set, constructs publishing content structure diagram, and in combination with publishing original resource set executes publishing field semantic information extraction and semantic correlation analysis, obtains publishing semantic diagram;From the structural distribution characteristics and content anchoring characteristics of semantic node in publishing resource in publishing semantic diagram are extracted, and the anchor point edge between publishing semantic diagram and publishing content structure diagram is established, and the multi-granularity anchor point double diagram is formed;Based on graph service execution plan, execute double diagram collaborative retrieval and service path generation, obtain knowledge service candidate set, and carry out service ordering and arrangement, generate publishing field knowledge service content.The application realizes the collaborative mapping of semantic relationship and content bearing position, enhances knowledge organization precision and service direction.

This invention discloses a method and apparatus for context optimization of a large-scale language model in the urban rail transit vertical domain based on dynamic cognitive graph construction and compression. The method first extracts a set of cognitive atoms from the urban rail transit domain text input in each round using the urban rail transit vertical domain large-scale language model. Then, based on the cognitive graph of the previous round and combined with the cognitive atom set of the current round, it completes node updates, relationship mining, and node cognitive activation strength calculation to obtain the current round's cognitive graph. Next, it calculates the cognitive saliency of each node, dynamically compresses the cognitive graph to obtain a compressed subgraph, compresses it into graph-aware prompt text containing node content and semantic relationships, and inputs it into the urban rail transit large-scale language model to generate a response. Finally, it extracts new cognitive atoms from the response to generate a supplementary atom set, analyzes the dependence of the response on nodes, and updates the model's attention. Both the supplementary atom set and the updated model attention provide support for the next round of cognitive graph updates. This invention solves the problems of invalid token accumulation and the inability of static context to dynamically focus and evolve in existing technologies.

The present disclosure sets forth systems, apparatuses, and methods that uniquely combine vector-based semantic processing with graph-based knowledge representations, leveraging the strengths of both approaches to improve similarity searches and context matching and capture nuanced semantic relationships between concepts. The disclosed systems, apparatuses, and methods include components configured to receive data, determine based on the data one or more entities, and generate, based on the one or more entities, a probability distribution, context associated with the data, and historical interactions, one or more semantic mappings comprising nodes interconnected according to weighted relationships, and update, based on at least one of additional data, context, or interactions, the one or more semantic mappings.

This invention discloses an information extraction method and apparatus for the maritime search and rescue field. The method includes: pre-setting a set of search and rescue case texts and a marine environmental spatiotemporal dataset; performing semantic information extraction processing on the search and rescue case text set to obtain a search and rescue case set; and constructing a search and rescue case atlas based on the marine environmental spatiotemporal dataset and the search and rescue case set. This invention can construct a maritime search and rescue case atlas with clear semantic relationships, unified description standards, and high completeness.

This invention relates to the field of resource management technology, and discloses a method and system for matching supply and demand of resources in the robotics industry based on knowledge graphs. The method includes: parsing resource data to obtain basic supply and demand data; semantically labeling the basic supply and demand data and mapping technical entities and indicator entities to a conceptual system to obtain a set of supply and demand capability features; using entities in the set of supply and demand capability features as resource nodes, writing nodes into a graph database to obtain resource entity nodes; performing link reasoning on the semantic relationships between resource entity nodes to obtain knowledge graph data; receiving matching requests, identifying the intent of the demand description, performing path retrieval in the knowledge graph data, and obtaining a preliminary matching result set; comprehensively sorting the supply entity nodes and encapsulating attribute information and associated paths into a protocol to obtain a matching scheme. This invention can improve the efficiency of supply and demand matching in the robotics industry.

The present application relates to the technical field of intelligent agricultural data platform task scheduling and cleaning, in particular to an agricultural heterogeneous big data platform adaptive task scheduling and data cleaning system. The present application establishes an index map of data and tasks, solves the correlation between task targets and task data through semantic analysis to establish a preliminary data fast call, establishes a complete task scheduling index through semantic relationship analysis of historical data tasks and task targets, filters the indexed data through automatic analysis and extraction of data cleaning rules, and dynamically adjusts the map during calculation according to the calculation resource occupation and data quality to accurately schedule and clean data according to task requirements, and adjusts the data in real time according to the task calculation resource condition.

The present disclosure relates to a data timeliness identification method and device, a content pushing method and device, an electronic device and a computer readable storage medium. The data timeliness identification method comprises: obtaining to-be-processed data; obtaining associated data by using a web crawler according to the to-be-processed data; and determining timeliness information of the to-be-processed data based on a semantic relationship between the to-be-processed data and the associated data, wherein the timeliness information comprises old news or non-old news. The real publication time of news content can be determined by using Internet information through the web crawler, the timeliness of the news content is further determined through the determined real publication time, whether the publication time of the news content is modified by a content cooperation party can be identified, and the news content in the database is marked with timeliness, so that the timeliness problem can be fully considered during subsequent pushing.

The application provides a city information model knowledge graph construction method and device and a readable storage medium. The method comprises: a CIM concept classification system: constructing conceptual knowledge as a standard for inducting or classifying specific CIM entities; CIM entity construction: converting discrete CIM data into specific objects according to the conceptual knowledge; CIM relationship extraction: extracting relationship information from various objects through the processes of knowledge extraction, relationship linking and relationship reasoning; CIM knowledge fusion: quantifying CIM knowledge into a CIM concept layer, a CIM entity layer and a CIM relationship layer by using a top-down construction method to obtain a CIM knowledge graph. The application combines actual CIM business needs, adds CIM business relationships on the basis of main semantic relationships, and is used for fusing semantic relationships and spatial relationships to realize large-scale multi-source heterogeneous data storage management and information retrieval.