1046 results about "Semantic relation" patented technology

The present invention relates to the technical field of big data processing, and discloses a multi-source heterogeneous data fusion and processing method based on big data, including the following steps: data preprocessing, cleaning, complementing and feature extraction of multi-source heterogeneous data, providing standardized and high-quality input for subsequent processing; dynamic knowledge graph modeling, constructing a dynamic knowledge graph based on the semantic relationship of multi-source data, and generating a unified semantic representation; cross-modal alignment and deep fusion, semantically aligning multi-modal data, and generating a unified comprehensive representation through deep fusion of features; real-time stream processing, dynamically updating and processing real-time updated stream data to achieve real-time fusion of data; adaptive optimization, dynamically adjusting fusion strategy parameters, and improving the adaptability and robustness of the fusion strategy. The present invention realizes efficient fusion, real-time processing and semantic consistency expression of multi-source heterogeneous data through dynamic knowledge graph modeling, cross-modal alignment, real-time stream processing and adaptive optimization.

The invention provides an AI intelligent customer service response method and system based on a remote digital service, and the method comprises the steps: firstly receiving a session request data flow inputted by a user through a remote session terminal, and extracting a dynamic semantic feature sequence containing a grammatical structure, an emotional tendency and a context dependence feature; inputting the data into an intention recognition model trained based on a historical session data multi-dimensional semantic relationship, and outputting an intention category label and confidence; matching a target dialogue stream node from a preset dialogue process library according to the label and the confidence coefficient; dynamic response content is generated based on the node response logic rule, and a response data stream is returned; and finally, collecting user feedback behavior data, updating the association weight of the intention recognition model according to an interaction validity index, optimizing the semantic mapping precision of the intention recognition model to similar session requests, and realizing more accurate, intelligent and adaptive customer service response.

The invention discloses a computer equipment fault monitoring system and method based on artificial intelligence, and relates to the technical field of computer equipment fault monitoring. The system comprises a data access module, a semantic analysis module, a knowledge graph construction module, a dynamic semantic association module, a data fusion processing module, a decision output module and an adaptive optimization module. The data access module collects and standardizes hardware, software and network data; the semantic analysis module extracts and enhances semantic tags; the knowledge graph construction module forms a data semantic relation network; the dynamic semantic association module screens potential semantic relationships; the data fusion processing module generates a multi-dimensional feature vector; the decision output module triggers fault early warning; and constructing a feedback knowledge graph of the self-adaptive optimization module. According to the method, through event-driven interpolation, dynamic weight fusion, closed-loop feedback optimization and the like, the problems of multi-source data alignment, semantic fusion and dynamic adaptation are solved, the fault monitoring accuracy and the system adaptability are improved, and the method is suitable for fault monitoring and early warning of computer equipment.

The invention discloses an intelligent file classification and retrieval method and system, and the method comprises the steps: carrying out the unified format conversion and feature extraction of text, image and audio data according to the original data format of a file through a multi-modal data fusion algorithm, and obtaining a standardized file feature data set; based on a deep learning adaptive classification model, performing classification processing on the standardized archive feature data set to generate multi-level archive classification labels; analyzing the semantic relationship between the archive classification labels through an enhanced semantic network construction technology, generating a semantic association map, and optimizing the semantic association map to obtain an optimized semantic association map; and according to a retrieval request input by a user, performing semantic matching and correlation sorting by utilizing the optimized semantic correlation graph, and outputting archive information most related to the retrieval request. By utilizing the embodiment of the invention, the relevance and accuracy of retrieval can be improved, and modernization and intellectualization of archive management are promoted.

The invention discloses a hybrid enhanced index and system based on vector retrieval and a BM2 algorithm, and the method comprises the steps: uploading a document by a user, analyzing the document by the system to obtain text content, segmenting the text content, converting the segmented text content into dense vectors, storing the dense vectors in a vector library, extracting keywords of segmented texts by using a large model, processing the keywords, and inserting the keywords into a word list. Word segmentation is carried out on the text based on the constructed word list and a word segmentation device; in response to the received user query, vectorizing the user query, and calculating the similarity with each vector in the vector library to obtain a preliminary query result; performing word segmentation on user query, performing keyword retrieval by using a BM25 algorithm according to a word list and a word segmentation device to obtain query results, and filtering the similarity of the two query results by the reordering model according to a set threshold value; selecting a result with the highest similarity according to large model parameter limitation and an upper limit set by business requirements; and splicing the document content corresponding to the result and the user query into a cue word, inputting the cue word into the large model, and analyzing a semantic relationship and a logic structure in the cue word to generate an answer.

The invention provides a streaming query semantic map adaptive enhancement method and system based on cognitive computing, and relates to the technical field of semantic maps, comprising the steps of computing node affinity through a map attention mechanism, identifying knowledge clusters by using a community discovery algorithm, performing knowledge deduction by taking the knowledge clusters as a basic unit, and predicting and complementing potential semantic relationships. And after multi-agent reinforcement learning optimization verification, outputting an enhanced semantic map and a credibility score table. And deploying the enhanced atlas to an edge network, carrying out parallel semantic understanding and query optimization suggestion generation, integrating optimization suggestions through federal learning, updating the central cognitive model, and distributing the central cognitive model to the edge network. According to the method, the semantic map is enhanced through cognitive calculation, so that more accurate semantic understanding and more efficient query optimization are realized.

The invention relates to the technical field of construction data modeling, in particular to a BIM (Building Information Modeling)-based construction progress dynamic optimization system, which comprises an abnormal event identification module used for detecting local abnormity based on multi-source data and marking an abnormal area; the local component reconstruction and dynamic fusion module is used for reconstructing an abnormal region component model and performing space mapping to obtain a global BIM model; the component-semantic double-layer graph modeling module is used for constructing a component topological dependency graph and a component semantic relation graph and jointly reasoning a state transition sequence of the component; the state recognition and self-adaptive classification module is used for fusing the geometric features of the component and the graph structure features and executing state recognition and judgment boundary adjustment; and the graph evolution calculation module is used for performing multi-round state propagation deduction to generate a component state evolution result. According to the method, in a heterogeneous information modeling environment, dynamic processing and state intelligent identification of component-level information are realized, and the data calculation precision and response capability under three-dimensional model driving are improved.

The invention belongs to the technical field of remote sensing image imaging, and provides a remote sensing image unsupervised domain adaptation method based on comparative learning and multi-prototype alignment, and the method specifically comprises the steps: constructing a teacher-student network, and setting a cross-domain multi-prototype alignment module and a pixel-level cross-domain comparative learning module which are needed in training in the teacher-student network; during teacher-student network training, a cross-domain multi-prototype alignment module is used for extracting prototypes of all categories in a source domain and a target domain and minimizing the feature distance of the prototypes, and a pixel-level cross-domain contrast learning module is used for explicitly modeling the global semantic relationship of cross-domain pixel samples; and an embedding space with high intra-class compactness and high inter-class separability is established.

The invention relates to the technical field of artificial intelligence, in particular to a multi-mode agent RAG-ReAct double-engine cooperative training method. The method comprises the following steps: acquiring multi-modal data, and converting the multi-modal data into a high-dimensional vector; constructing a knowledge graph based on the high-dimensional vector; encoding the semantic relationship of the knowledge graph into a model fine adjustment gradient direction by using a dynamic distillation technology; designing a distributed architecture based on the high-dimensional vector; performing hybrid retrieval based on a distributed architecture to obtain hybrid retrieval data; executing distributed reasoning based on a preset recursive reflection mechanism and the mixed retrieval data to generate distributed reasoning data; performing data parallel detection according to the distributed reasoning data to obtain data parallel parameters; and performing node resource scheduling based on the data parallel parameters so as to obtain node load balancing data. Based on the artificial intelligence technology, the reasoning accuracy and the resource utilization rate of the multi-modal agent in a complex task environment are effectively improved.

The invention discloses a knowledge conversion and fusion processing method and system for massive power grid operation data, and the method specifically comprises the steps: obtaining the attributes, including the alarm type, timestamp and severity, of equipment granularity based on alarm event data collected by a power distribution automation terminal, and carrying out the data cleaning and standardization processing, converting the multi-source heterogeneous alarm data into uniform event granularity representation; fusing the alarm event data and the operation event data, carrying out space-time alignment and causal association on the two types of events based on timestamps and equipment ID attributes, constructing a cross-data-source event semantic link network, and carrying out association analysis on a semantic relationship between the events; and according to a fault diagnosis result, historical records related to fault processing are extracted from the operation event knowledge base, similar fault processing modes are matched, and a specific fault processing scheme is generated in combination with factors of severity and influence range of the current fault.

The invention relates to the technical field of medical data management, in particular to a medical data document classification and marking system which comprises a document structure analysis module, a semantic information identification module, a classification and marking module and a document intelligent label module. According to the method, document visualization and retrieval efficiency is improved by analyzing document content and extracting titles, texts and appendixes, analytic performance and operability of data are enhanced by identifying key information and semantic relations and accurately extracting and labeling medical terms, medicine research data and medical record records, storage is optimized according to priorities in combination with classification and marking, and the method is high in practicability and easy to popularize and use. Clinical values and application scenes of documents are clarified, document management processes are refined, management compatible with clinical data formats is formed through structured storage of marked data and classification label generation, storage normalization and retrieval efficiency are improved, the requirement for accurate information calling in clinical tests is met, document management capacity is improved, and the document management efficiency is improved. And support is provided for improvement of clinical decision-making and research efficiency.

The invention discloses a cockpit active recommendation system and method based on a knowledge graph and semantic reasoning, and relates to the technical field of intelligent cockpits. The system receives natural language voice input of a user, executes voice recognition and semantic analysis, extracts user intention, keywords and slot entities, generates structured semantic information, constructs or calls a knowledge graph structure with semantic relation edges in combination with environment context information, and obtains the knowledge graph structure with the semantic relation edges. Semantic path reasoning is carried out based on the path dependence weight and the semantic similarity, a semantic edge label guided graph attention mechanism is introduced to calculate a path consistency score, a candidate recommendation set is generated, the semantic fitting degree and the path score are fused to sort and output recommendation content, and the graph edge weight and the user portrait are updated based on user feedback. According to the method, semantic understanding precision, recommendation path interpretability and system adaptive capacity are improved, and the method is suitable for personalized voice recommendation, man-machine interaction and scene linkage control tasks in an intelligent cockpit.

The invention relates to the technical field of artificial intelligence and big data analysis, in particular to a data operation system and method based on a knowledge graph, and the method comprises the steps: extracting an entity and semantic relationship from multi-source business data, and constructing a dynamic evolvable initial knowledge graph; node features are aggregated, and multi-dimensional situation state vectors are generated in combination with gating loop unit modeling behavior path dependence; through a structure-semantic coupling attribution scoring mechanism, a statistical information gain and semantic similarity are fused to identify a core driving factor, and a causal regression model of the factor and an operation target is established; dynamically adjusting the edge weight and the structure of the atlas in real time, and triggering a new path discovery mechanism to continuously optimize the atlas; and according to a quantitative business target, reversely extracting a high-confidence influence path from the atlas, and generating a personalized strategy combination through intervention simulation and multi-target Pareto optimization, thereby realizing intelligent recommendation and decision closed loop driven by an operation target. According to the invention, higher-precision operation situation awareness and strategy generation are realized.

The invention relates to a government affair file information extraction and question answering method and device and a medium, and the method comprises the steps: carrying out the entity extraction of a government affair file through employing a BERT-CRF joint model, and obtaining a structured entity set; performing relation extraction on the structured entity set to generate a semantic relation set between the entities; constructing a knowledge graph according to the structured entity set and the semantic relationship set, storing entity nodes into a graph database, and storing an embedded vector of an entity text into a vector database; when a query request of a user is received, relation query of the graph database and semantic retrieval of the vector database are carried out, sub-graph structures and semantic matching vectors related to query are extracted, and a mixed retrieval result is obtained; and inputting the mixed retrieval result into a large language model, and generating a question and answer response text conforming to a preset format by applying a dynamic prompt template. According to the method, the document processing efficiency and accuracy are effectively improved, and a solid technical support is provided for intelligent management of government affair documents.

The invention provides a semantic enhancement and dynamic completion method and system for a power market data graph. The method comprises the following steps: acquiring target entity information and external data of a power market data graph by responding to a data query instruction; converting external data into structured semantic tags, then traversing the topological structure, establishing a mapping relationship between the tags and entity attributes, and combining query instruction influence logic to calculate topological integrity indexes and mark missing entities or correlate to form a missing tag set; establishing attribute mapping based on a target entity and external data, generating a semantic vector fusing cross-domain features, and projecting the semantic vector to a topological space according to a missing mark set to form cross-domain feature topological nodes and connections; embedding a missing position to reconstruct a topological structure, and checking logic connectivity; the dynamic conduction chain is analyzed, and a feedback interface containing the semantic relation and the completion structure is generated to respond to query. According to the method, conduction logic closed-loop verification of the missing elements in the power market data graph is realized.

The invention discloses a multi-modal large model optimization method and system based on defect detection and analysis. The method comprises the following steps: S1, collecting a multi-modal original sample with power business scene characteristics; s2, performing standardization preprocessing on the original data to obtain a fine-tuned standard data set; and S3, establishing relational mapping among the labels of the training data set to form a semantic relation graph of the labels, and extracting a topological structure relation of the semantic relation graph to generate the power text knowledge base. S4, various task function modules of the computer vision large model are trained, so that the computer vision large model has various tasks for processing the electric power scene. And S5, fusing the trained Clip semantic large model and the trained computer vision large model into a multi-modal large model by using a Transform structure. According to the method, the multi-modal large model subjected to fine tuning through the power text knowledge base not only maintains the specific analysis and reasoning performance under various power scenes, but also can ensure the generalization ability of the large model to various algorithm tasks.

The invention discloses a unified management method and device for multi-protocol Internet of Things equipment, and relates to the technical field of Internet of Things communication and protocol management, and the method comprises the steps: constructing a cross-protocol semantic knowledge graph through a graph neural network based on a structured gene-semantic mapping table, and generating a dynamic semantic relation matrix through a node embedding method; according to the dynamic semantic relationship matrix, deploying a genetic algorithm engine at an edge node to dynamically recombine and optimize protocol genes, screening protocol variants adapted to a network environment through an evolution sandbox pressure test, and generating an optimized protocol gene pool; and based on the optimized protocol gene pool, performing real-time conversion on the heterogeneous protocol message in a protocol conversion gateway in combination with a multi-thread parallel analysis architecture to generate a standardized semantic data stream. Unified expression and management of complex heterogeneous protocols are achieved, the protocol conversion process is simplified, cross-protocol semantic understanding and compatibility are improved, and device flexibility and analysis efficiency are improved.

A system and method for latent hyperspace navigation in spatiotemporal media using hierarchical and Lorentzian autoencoders. The system compresses spatiotemporal media into navigable latent representations while preserving geometric and semantic relationships through tensor structure maintenance. A latent hyperspace manager organizes compressed representations as geodesic trajectories within a geometric manifold structure based on differential geometry principles. A geodesic trajectory mapper computes optimal navigation paths through the high-dimensional space, while symbolic anchors positioned at semantically significant locations serve as persistent reference points. Spatiotemporal routing protocols manage navigation decisions across multiple temporal scales. A strategy caching system preserves successful navigation patterns for reuse, enabling continuous learning. The system generates synthetic content during navigation to support infinite zoom capability, allowing exploration beyond original media boundaries. Cross-modal fusion combines diverse input modalities into unified representations, applicable to immersive media exploration, scientific visualization, and surveillance analysis.

The invention discloses a semantic segmentation method for a low-resolution road scene, and aims to solve the problems of difficulty in small target recognition, fuzzy details, texture information loss and the like existing in a low-resolution image in the conventional semantic segmentation technology. The method comprises the following steps: (1) collecting a low-resolution road scene image and a corresponding semantic tag; (2) constructing a semantic segmentation model consisting of an edge guidance module (BGM), a double-domain feature decomposer (DDFD), a domain alignment attention fusion module (DAAFM) and a double-layer attention context aggregation module (HACAM); (3) designing a joint loss function to carry out multi-scale supervision on semantic regions, edges and middle features; (4) carrying out model training by utilizing the road scene image; and (5) outputting a semantic segmentation result map and an edge prediction map. The boundary perception capability is enhanced by introducing learnable pixel difference convolution, the extraction precision of a small target and a global structure is improved by combining frequency domain and spatial domain feature alignment, and context semantic relationship expression is optimized by fusing a channel and a spatial attention mechanism. The method effectively improves the semantic segmentation precision and boundary restoration capability of the model in a low-resolution complex road environment, and is suitable for intelligent analysis tasks of road images in scenes of automatic driving, intelligent traffic, severe weather and the like.

The invention relates to the technical field of intelligent document processing, and particularly discloses a file arrangement processing method and system. The method comprises the following steps: acquiring an input document, analyzing the structure and the style of the input document, and generating standardized format data; extracting a semantic relationship of the document through a pre-trained text understanding model, constructing a semantic relationship graph and optimizing a chapter sequence; loading a document processing plug-in to execute cross-format arrangement, and generating intermediate format data; executing processing nodes based on a task scheduling mechanism to generate optimized content; calling a multi-format output module to generate a target document, and adjusting metadata in combination with the global variable; and finally evaluating the document quality and updating the model. According to the method, automatic analysis, intelligent optimization and format unification of the multi-source heterogeneous document are realized, and the intelligent degree and the output efficiency of document processing are improved.

The invention relates to the technical field of synchronous positioning and map construction, in particular to a geometric-semantic collaborative fusion mobile robot three-dimensional semantic map construction method, which is used for performing deep fusion on a geometric map and multi-view semantic information so as to solve the problem of difference of the two types of information in resolution, scale and dimension. The geometric map mainly provides structured information of the environment, and the semantic information describes object categories and features thereof in the environment. Direct fusion of the two types of information often causes spatial positioning errors and semantic information loss or redundancy problems. According to the method, a geometric map containing environmental structured information is constructed through an SLAM technology, and meanwhile, object classification, attribute labeling and semantic relationship analysis are performed on multi-view image data by utilizing the semantic information extraction capability of a large model. And finally, precise fusion of the geometric map and multi-view semantic information is realized, and meanwhile, the object recognition and semantic understanding capabilities of the robot in a complex scene are improved.

The invention discloses a dialogue recommendation method based on combination of a multi-modal semantic graph and prompt learning, which comprises the following steps of: firstly, coding entities in dialogue history based on a knowledge graph, carrying out feature coding on a product text and an image, and then, constructing a plurality of semantic graphs with different modals, including a text semantic graph, an image semantic graph and a collaborative semantic graph, on the basis, a multi-modal semantic graph is innovatively combined with prompt learning, high-dimensional semantic association is fully mined by using a large language model, multi-modal features and structured semantic relationships between products are effectively modeled, and the multi-modal semantic graph is embedded into the multi-modal semantic graph embedded # imgabs0 # with multiple modal semantic relationships fused therein, so that the multi-modal semantic graph embedded # imgabs0 # and the multi-modal semantic relationships between the products are effectively modeled, and the multi-modal semantic graph embedded # imgabs0 # and the multi-modal semantic relationships between the products are obtained. The user preference is dynamically captured in the multi-round dialogue recommendation scene to comprehensively understand the user demand and generate a more accurate recommendation result, so that the accuracy in the recommendation task is remarkably improved, meanwhile, more natural content conforming to the context can be generated in the dialogue task, and the dialogue generation performance of the dialogue recommendation system is improved.

The invention relates to the field of agricultural information processing and artificial intelligence, in particular to a farmland yield prediction method and system based on a heterogeneous graph neural network, and the method comprises the steps: obtaining multi-source farmland data, and extracting an initial feature vector of a farmland plot node; on the basis of the initial feature vector, constructing a heterogeneous graph structure containing multiple semantic relationships; performing node feature updating on the heterogeneous graph structure by using a heterogeneous graph neural network, and dynamically aggregating information of multiple types of neighbor nodes through relation-aware message passing and an edge propagation gating mechanism; performing enhancement processing on the node features by using a space-time dependency enhancement mechanism and a knowledge-guided reasoning mechanism; and outputting a regression prediction result of the farmland yield through a prediction module based on the enhanced node features. The invention aims to realize modeling and intelligent yield prediction based on multi-source heterogeneous data in an agricultural system, and improve the environmental adaptability, prediction generalization ability and interpretability of farmland yield prediction.

The invention discloses a document link construction and evolution relation tracking method based on fragment-level semantic alignment, which relates to the technical field of document management, and comprises the following steps: carrying out structured analysis on a plurality of input documents, and dividing the document content into a plurality of semantic fragments; performing vectorization processing on each semantic fragment, and constructing a document-fragment-vector semantic mapping relationship; different document fragment-level semantic alignment is realized through cross-document semantic vector similarity calculation, a position offset tolerance algorithm and context association analysis, and a semantic link network is established; fusing the link relationship, the timestamp, the version information and the author information, and constructing a document evolution diagram taking the documents or fragments as nodes and evolution links as edges to record semantic flow paths between the documents; and outputting the semantic link network and the document evolution diagram in a structured form. The method can solve the problems that semantic relations between documents are missing, evolution paths are invisible, content reuse cannot be traced, and the structure induction capability is insufficient.

The invention discloses a multi-modal metadata alignment fusion method and device, equipment and a storage medium, and the method comprises the steps: carrying out the metadata extraction of structured data, unstructured text data and image data, and generating multi-modal metadata with semantic annotations; establishing a shared semantic embedding space, and mapping the multi-modal metadata to the shared semantic embedding space for coding to obtain a unified spatial vector; determining an alignment candidate pair from the unified spatial vector through similarity calculation, and identifying a semantic relationship of the alignment candidate pair; and performing conflict detection on the aligned candidate pairs and the corresponding semantic relationships, and resolving conflicts based on weight weighting fusion to obtain a unified metadata system. According to the method, improvement and optimization are carried out from multiple aspects of multi-modal data processing, semantic understanding, alignment accuracy, conflict resolution and the like, the defects in the prior art are overcome, and a more accurate and comprehensive multi-modal metadata alignment fusion result can be provided.

The invention provides a structured processing method and device for a contract text, equipment and a storage medium, and the method comprises the steps: carrying out the preprocessing of the contract text, generating a standardized word sequence, and enabling the standardized word sequence to comprise a plurality of word sequences and part-of-speech tagging information of each word sequence; the standardized word sequence is input into a named entity recognition model to recognize key entities in the contract text, key terms in the contract text are extracted based on a machine learning algorithm, the key entities include both parties of the contract, the amount of the contract and the name of a product, and the key terms include payment terms, delivery terms, default responsibility terms and force majeure terms; through dependency syntactic analysis and a semantic role labeling model, analyzing to obtain a semantic relationship between the key entities and the key terms; and storing the extracted key entities, key terms and semantic relationships in a target database in a structured form, and establishing a target index to support quick query. According to the invention, the error rate of manual processing is reduced.

The invention provides a method and a system for intelligently generating and dynamically adjusting a public emergency plan, which are oriented to the field of public safety emergency. The method comprises the following steps: fusing multi-source heterogeneous data to construct an extensible domain knowledge graph, establishing a standardized emergency instruction library, and carrying out multi-dimensional information labeling; automatic extraction of disaster elements is realized based on an entity recognition model of deep learning; analyzing association rules among the emergency entities through a semantic relationship mining technology; key information such as a disaster chain and resource distribution is rapidly obtained by using a map reasoning mechanism; carrying out cross-department resource collaborative allocation by adopting a multi-objective optimization algorithm to generate an optimal disposal scheme; a high-dynamic adjustment mechanism is constructed, and an emergency plan is dynamically optimized based on situation evolution prediction and real-time monitoring data; and finally, automatic generation and versioning management of the plan are realized. Through multi-mechanism cooperation of knowledge modeling, intelligent element analysis, semantic reasoning, dynamic optimization and predictive adjustment, the emergency plan generation efficiency, situation adaptability and resource allocation rationality are remarkably improved, and support is provided for quick response and scientific decision under emergencies.

The invention discloses a cross-modal image text retrieval method based on deep learning, and provides a novel cross-modal feature extraction and alignment framework, efficient cross-modal retrieval is performed by learning semantic representation of an image and a text in a common feature space, the framework comprises a plurality of modules, and in a feature extraction stage, the cross-modal feature extraction and alignment framework is used for extracting and aligning the images and the texts. BERT Tokenizer and BERT Embedding are adopted for text coding, and the image feature extraction is combined with a Faster R-CNN (Convolutional Neural Network) model and a ResNet-101 model, so that the dimensional consistency of image features and text features is ensured. In the feature alignment stage, fine-grained semantic alignment of image and text features is realized through an I2T Attention module and a T2I Attention module, and the cross-modal matching precision is remarkably improved. In the correlation scoring stage, a correlation scoring matrix is generated by calculating the similarity score between the image features and the text features, so that the accuracy of cross-modal matching is improved. And further optimizing the feature alignment effect through normalization and attention redistribution. In an optimization stage, the system innovatively adopts a shared semantic and sorting loss strategy, integrates an OpenCLIP model framework, efficiently mines a semantic relationship between an image and a text from unlabeled large-scale data through comparative learning, and shows strong transfer learning ability and inference precision.

The invention provides a government affair hotline service knowledge graph construction method and system based on a large language model, and relates to the technical field of natural language processing and government affair informationization, and the method comprises the steps: obtaining a government affair hotline service scene, and defining a mode layer of a multi-scene knowledge graph; unstructured government affair hotline question and answer data are collected, semantic fusion, logic extraction and ambiguity elimination are carried out on the data based on a chained thinking mechanism of a large language model, and a fused government affair hotline question and answer set is generated; driving a large language model to identify knowledge entities in the set by adopting a step-by-step prompt strategy, extracting semantic relationships among the knowledge entities by adopting a semantic association analysis method, constructing triple knowledge data based on the knowledge entities and the semantic relationships, storing the triple knowledge data into a graph database, and performing dynamic updating and multi-dimensional retrieval on the multi-scene knowledge graph to obtain a multi-scene knowledge graph set; therefore, a large language model and a knowledge graph technology can be fused, government affair data are efficiently integrated, and the intelligent level of government affair service is improved.

The invention discloses a method for converting a natural language into a low code based on multi-modal reinforcement learning, which comprises the following steps of: performing word segmentation, embedding and multi-layer feature extraction on a natural language instruction input by a user, combining a self-attention mechanism and a graph attention mechanism, extracting and optimizing an original semantic feature vector, and automatically identifying a business field; semantic relationship triples in the domain knowledge graph are fused, and a multi-modal semantic alignment and context enhancement strategy is adopted, so that the accuracy and stability of semantic representation are remarkably improved; when semantic drift or ambiguity is detected, a multi-candidate correction mechanism is triggered to obtain a better analysis result, and the robustness of semantic understanding is enhanced; and finally, mapping an analysis result into an instruction which can be identified by a low-code platform, automatically generating a code structure, continuously optimizing a semantic model and a knowledge graph based on user feedback, realizing adaptive learning, and improving the conversion efficiency and quality from a natural language to codes.