22 results about "Semantic link" patented technology

The invention provides a context compression method based on multi-round dialogue intention graph construction, which comprises the following steps: S1, dialogue data acquisition and preprocessing: carrying out natural language processing on each dialogue unit; s2, intention atlas construction is achieved through node design and edge design, and each node comprises original text content and structured semantic information; s3, carrying out context compression and graph structure cutting, and only retaining sub-graphs forming a core semantic link; s4, dynamic context management and topic jump processing: in a multi-topic dialogue, when a user jumps or switches to a new topic, a system records a sub-graph of a current active topic, and contextual nodes of an inactive topic are frozen; s5, context sequence generation and model input: linearizing node contents in the cut sub-graph according to a dependent link sequence to generate a compressed context sequence, and transmitting the context sequence and current user input to a large language model for reasoning; and S6, continuous updating and feedback optimization are carried out.

The invention discloses a flexible operation and maintenance early warning device based on artificial intelligence. The flexible operation and maintenance early warning device comprises a multi-source data acquisition module, a data preprocessing and fusion module, an AI state recognition module, a risk reasoning and knowledge graph module, a flexible strategy engine module and an early warning output module. The device realizes cleaning, noise reduction, time alignment and feature fusion of multi-source data by collecting equipment operation indexes, log behaviors, security events and environmental parameters. The AI state recognition module is used for recognizing an operation state, a behavior mode and potential abnormity, and the risk reasoning and knowledge graph module completes risk source positioning and reason verification based on an entity relationship and a semantic link. And the flexible strategy engine dynamically generates an early warning strategy according to a reasoning result, and pushes early warning information in a multi-channel manner through an early warning output module. The device can realize high-accuracy, interpretable and self-adaptive operation and maintenance early warning capability, and is suitable for intelligent operation and maintenance management of hospital machine rooms and key business systems.

The invention relates to the technical field of semantic retrieval, in particular to an intelligent law retrieval system based on a semantic reasoning atlas, which comprises a semantic element extraction module for extracting law elements to generate a semantic unit set, a law semantic atlas construction module for establishing concept nodes and reasoning atlases, a semantic association reasoning module for screening paths to generate a reasoning sequence, and a semantic association reasoning module for establishing a semantic association reasoning sequence. The retrieval intention mapping module constructs a mapping set in combination with user query, and the legal result aggregation module matches provisions and cases and generates legal semantic retrieval results through screening, sorting and integration. According to the method, law element dependency is recognized through semantic decomposition, a causal semantic link is formed through structured aggregation expression, logic paths are dynamically integrated to achieve law association accurate modeling, query intention strong mapping is constructed through semantic similarity and logic constraints, the result interpretation depth and aggregation capability are improved, matching is made to conform to the law context, and the method has the advantages of being high in practicability and easy to popularize. And the identification and inference precision of the system on logic association is effectively enhanced.

In an illustrative embodiment, systems and methods for extracting, and organizing, and visualizing details of options provided by multiple organizations responsive to a risk fulfillment request include analyzing unstructured electronic documents to recognize various quote aspects in their contents, label the quote aspects according to a classification, and store semantically linked quote aspects. The systems and methods, for example, may enhance semantically-linked groups of quote aspects with attributes according to a corresponding ontology, and confirm the labeling, grouping, and enhancing through feedback interactions performed with a user via a graphical display. The confirmed information may be used to generate a visualization of options for fulfilling the request, each option qualified and / or color-coded through automated learned analysis for review by the user.

The application discloses an improved knowledge graph construction method and system for textile weaving, relates to the technical field of textile manufacturing, and solves the problems of insufficient feature description ability of the knowledge graph of the prior art for the textile weaving industry, difficulty in expressing the time sequence logic evolution characteristics and the causal relationship under the multi-source disturbance, lack of entity boundary definition and semantic recognition mechanism for weaving professional terms, and serious semantic mismatch of relationship extraction, the method comprising: constructing a knowledge base data set of the improved knowledge graph, modeling five concept domain ontologies, defining a semantic link paradigm, enhancing co-occurrence of graph entity boundaries, and enhancing attention graph extraction with relationship type perception, carrying out weaving knowledge entity recognition and relationship extraction result reasoning output, introducing a structured energy consumption semantic network to realize the transition of the model from data driving to knowledge driving, and realizing the transition from "numerical correlation" to "causal interpretability" through an embedded semantic layer fusion mechanism.

The invention relates to the technical field of enterprises, and aims to solve the problems that in the existing expert semantic network construction process, an expert still stays in an unstructured carrier of text archives, manual records and experience summarization, so that the ability boundary of an expert is fuzzy, professional attributes lack clear semantic connection, and systematic semantic link expression is difficult to form; the enterprise field expert semantic network construction system comprises a pre-division module, an updating module, a reconstruction module and a network construction module. Through semantic hierarchy construction, relation extraction and network expansion mechanisms, the internal logic of expert knowledge can be subjected to deeper semantic modeling, and the expert knowledge originally dispersed in documents, experience records and unstructured texts can be extracted, analyzed and organized in a unified manner, so that the expert knowledge has expandability, relevance and reasonability, and the expert knowledge can be extracted, analyzed and organized in a unified manner. Therefore, the integrity and accuracy of the expert knowledge in the enterprise field in the semantic description level are effectively improved.

The invention discloses a differentiated building efficient interaction sensing general calculation modular intelligent connection regulation and control device and application, and belongs to the technical field of building electric power regulation and control. The device comprises a sensing unit, a communication unit, a computing unit and a cloud platform which are connected through a unified bus. The sensing unit collects multi-dimensional information and guarantees data quality, the communication unit integrates multiple protocols to achieve cross-protocol semantic intercommunication, the computing unit architecture adapts to different requirements of differentiated buildings, and the cloud platform provides management scheduling services. The invention further provides a hierarchical and domain-divided building power grid interaction semantic modeling method, three layers of sub-models and an end-to-end semantic link are constructed, and efficient interaction between the building and the power grid is achieved. The method solves the problem of pain points such as heterogeneous intercommunication, adapts to various types of buildings, and provides support for transformation of the buildings to active energy units and construction of novel electric power systems.

The invention provides a bidding file abstract generation method and system based on a large language model, and relates to the technical field of natural language process.The method comprises the steps that firstly, domain terms are extracted from an input to-be-processed bidding file set to construct a term library, and file units are split to form a domain semantic chain; inputting the term library into a pre-trained large language model to adjust the semantic analysis weight so as to obtain a semantic analysis model adaptive to the field; inputting semantic nodes in the semantic chain into the semantic analysis model to extract core expressions so as to form a dynamic abstract fragment set; determining an abstract fragment association sequence according to a semantic node connection relationship, and joining to obtain a preliminary fusion abstract text; and finally, semantic feedback of the reader is received, abstract fragment expressions and sequences are adjusted, a semantic chain is updated, and a final abstract is generated and output, so that the high-quality bidding document abstract can be efficiently generated.

The invention discloses a quota term content standardization conversion method based on semantic link and retrieval enhancement, and relates to the technical field of standardization conversion, the method comprises the following steps: determining a structured field of a quota term based on multi-source data of power grid quota business; performing semantic understanding on a to-be-processed quota clause text through the trained large language model, mapping a corresponding field in the quota clause text into a structured field, and outputting the structured field; performing grammar and business verification on the structured field output by the large language model, and obtaining an object candidate field from the verified structured field; performing standard entity alignment in a pre-constructed power grid entity dictionary and a pre-constructed knowledge graph through a semantic link module according to the object candidate fields; inputting the structured field aligned by the standard entity into a retrieval enhancement generation module used for realizing structured quota record joint generation; the whole process is well documented and is convenient to audit and maintain.

The invention provides a traffic logistics-oriented large model knowledge base construction method and system, and the method comprises the steps: carrying out the semantic relation mining based on a standard term set in the traffic logistics field, generating a structured term graph, constructing a template knowledge probe set, inputting a pre-training large language model, and carrying out the directional activation detection operation, thereby obtaining an implicit knowledge neuron cluster, the method comprises the following steps of: carrying out space-time correlation analysis on an activation response mode of the neural network, extracting distribution characteristics of neuron activation intensity and a dependency relationship among the characteristics, mapping an implicit knowledge neuron cluster into a semantic traffic logistics field knowledge unit, carrying out hierarchical organization and semantic linking according to a logic correlation degree and a functional attribute classification result, and carrying out hierarchical classification on the neural network. Generating a traffic logistics field structured knowledge network; and calling a standard traffic logistics business problem set to perform knowledge utility verification, and dynamically optimizing and adjusting knowledge units and association relationships to obtain a large model knowledge base, thereby providing high-quality knowledge support for intelligent application in the field of traffic logistics.

The application provides a semantic task processing method and device, electronic equipment and storage medium. The semantic task processing method comprises: acquiring language interaction information input by a user in a multi-round dialogue, analyzing the language interaction information, and generating a plurality of basic semantic tasks; generating a semantic task link based on a dependency relationship between the basic semantic tasks, and assigning a corresponding task intelligent agent to each basic semantic task based on the semantic task link; executing the corresponding basic semantic task through each task intelligent agent, generating a task signature of each basic semantic task, and generating a semantic link fingerprint based on the task signature. The application analyzes the language interaction information into basic semantic tasks, structures the basic semantic tasks into executable and traceable semantic task links, and finally generates a semantic link fingerprint, which guarantees the flexibility of the generated question and answer, and improves the controllability, reliability and traceability of complex multi-step queries.

In an illustrative embodiment, systems and methods for extracting, and organizing, and visualizing details of options provided by multiple organizations responsive to a risk fulfillment request include analyzing unstructured electronic documents to recognize various quote aspects in their contents, label the quote aspects according to a classification, and store semantically linked quote aspects. The systems and methods, for example, may enhance semantically-linked groups of quote aspects with attributes according to a corresponding ontology, and confirm the labeling, grouping, and enhancing through feedback interactions performed with a user via a graphical display. The confirmed information may be used to generate a visualization of options for fulfilling the request, each option qualified and / or color-coded through automated learned analysis for review by the user.

The application provides a bidding document abstract generation method and system based on a large language model, and relates to the technical field of natural language processing. First, the inputted to-be-processed bidding document set is extracted to construct a term library, and the document unit is split to form a domain semantic chain. Then, the term library is inputted into a pre-trained large language model to adjust the semantic parsing weight, so as to obtain a domain-adapted semantic parsing model. Then, the semantic nodes in the semantic chain are inputted into the semantic parsing model to extract core expressions and form a dynamic abstract fragment set. Then, the abstract fragment association order is determined according to the connection relationship of the semantic nodes, and a preliminary fusion abstract text is obtained. Finally, the reader semantic feedback is received to adjust the abstract fragment expression and order, the semantic chain is updated to generate a final abstract, and the final abstract is outputted, so that a high-quality bidding document abstract can be efficiently generated.

The invention discloses a primary school mathematics multi-modal mixed reality learning resource generation method oriented to body agent teaching, and the method comprises the steps: mapping a symbolized entity in a question into a physical attribute of a body environment through mathematical entity dynamic modeling based on a physical property role theory, and achieving the semantic linkage of a symbol system and an entity environment; the problem solving step is dynamically converted into a multi-modal resource through a guided dialogue driven by a physical relation and generation of a situational schematic diagram, the generated three-dimensional mathematical entity is dynamically embedded into a physical environment through a teaching scene modeling and adaptation mechanism based on Gaussian splashing, and an intelligent agent is supported to guide a learner in an MR environment. And dynamically optimizing learning resources and spatial layout through a task-scene joint representation model, so that the teaching sequence is adaptive to the cognitive state and environmental characteristics of the learner. Therefore, a dynamically interactive mixed reality learning environment is provided, learners are guided to complete mathematical tasks in real time based on self-generated resources, and efficient man-machine collaborative teaching is realized.

The present application relates to a large model intelligent agent Prompt optimization method, system and device, belonging to the field of artificial intelligence. The method comprises: receiving an initial Prompt and extracting semantic features; intelligently recommending optimization templates, frameworks or models from a resource library based on the features; generating an optimized Prompt using the recommended resources; calling multiple large model intelligent agents to actively evaluate the Prompt and generate feedback suggestion labels; and iteratively optimizing based on user feedback. The system includes four core modules: input processing and version management, intelligent recommendation, generation optimization, feedback analysis and active mining. The present application realizes the tracing and management of the optimization process by constructing a version evolution graph that records the optimization semantic link, accurately matches optimization resources through intelligent recommendation driven by semantic features, and effectively solves the problems of chaotic Prompt optimization process, low efficiency and dependence on expert experience in the prior art through fuzzy feedback analysis and multi-agent active evaluation optimization of human-computer collaboration, thereby improving the automation level and quality of Prompt optimization.

The application relates to the field of artificial intelligence technology and discloses a predictive dialing and multi-round dialogue system of an intelligent outbound robot, which comprises a business subdomain module, a business label generation module, a business template resetting module and a dialogue record generation module. The business subdomain module is used for obtaining an outbound task number set and dividing each business in the outbound task number set into multiple business domains. The business label generation module is used for triggering the dialing behavior of each line according to a business domain dialing scheduling set and writing the dialing behavior of each line into a corresponding business domain label to generate a line business label. The business template resetting module is used for determining a business dialogue template based on the line business label corresponding to each line, executing a template resetting mechanism and obtaining a reset business dialogue template. The dialogue record generation module is used for executing closed-loop processing of corresponding businesses for each line according to the reset business dialogue template, archiving the current business dialogue to the corresponding business domain and obtaining a business domain dialogue record set, so that the business semantic link of each outbound line remains independent.

The application provides a large model knowledge base construction method and system for traffic logistics, based on the standard term set of the traffic logistics field, semantic relationship mining is performed to generate a structured term atlas, a template knowledge probe set is constructed, a pre-trained large language model is input to perform directional activation detection operation, and an implicit knowledge neuron cluster is obtained, a spatiotemporal correlation analysis is performed on the activation response mode thereof, the distribution characteristics of the neuron activation intensity and the dependency relationship between the characteristics are extracted, the implicit knowledge neuron cluster is mapped to a semantic traffic logistics field knowledge unit, hierarchical organization and semantic linking are performed according to the logical correlation degree and the functional attribute classification result, and a structured knowledge network in the traffic logistics field is generated; a standard traffic logistics business problem set is called to verify the knowledge utility, the knowledge unit and the associated relationship are dynamically optimized and adjusted, a large model knowledge base is obtained, and high-quality knowledge support is provided for intelligent application in the traffic logistics field.

The application discloses a primary school mathematics multi-modal mixed reality learning resource generation method for embodied agent teaching, and realizes semantic linking of a symbol system and an entity environment by mapping a symbolic entity in a question to a physical attribute of an embodied environment through dynamic modeling of a mathematical entity based on a material property character theory; a problem solving step is dynamically converted into multi-modal resources through guided dialogue and situational sketch generation driven by a material property relationship, three-dimensional mathematical entities generated are dynamically embedded in a physical environment through teaching scene modeling and an adaptive mechanism based on Gaussian splashing, and an agent supports a learner in a MR environment. Through a task-scene joint representation model, learning resources and spatial layout are dynamically optimized, and a teaching sequence is adapted to a cognitive state of the learner and environmental characteristics. Thus, a dynamically interactive mixed reality learning environment is provided, and the learner is guided in real time to complete a mathematical task based on self-generated resources, and efficient teaching of human-computer collaboration is realized.

The application discloses a kind of behavior analysis method and system based on image recognition, comprising the following steps: obtaining and preprocessing continuous video stream data in target scene;Target detection and cross-frame association processing are executed in the image sequence to be analyzed;Multi-dimensional image recognition features are extracted in continuous tracking trajectory sequence;Behavior semantic unit is constructed, and semantic state sequence is generated according to the time sequence of each behavior semantic unit;Behavior segment division is carried out on target object based on semantic state sequence;Behavior semantic analysis link is constructed based on behavior segment sequence, and behavior evolution feature sequence is generated based on behavior semantic analysis link;Improved PoseC3D model is used to perform behavior recognition processing, and behavior analysis result is generated.The application realizes continuous behavior analysis by multi-dimensional feature fusion and semantic link modeling method, with the advantages of high recognition accuracy and strong abnormal detection capability.

The invention relates to the technical field of artificial intelligence, and discloses a predictive dialing and multi-round dialogue system of an intelligent outbound robot, and the system comprises a service domain division module which is used for obtaining an outbound task number set, and dividing each service in the outbound task number set into a plurality of service domains; and the service label generation module is used for triggering a dialing behavior of each line according to the service domain dialing scheduling set, and writing a corresponding service domain label into the dialing behavior of each line so as to generate a line service label. And the service template resetting module is used for determining a service dialogue template based on the line service label corresponding to each line, and executing a template resetting mechanism to obtain a reset service dialogue template. And the dialogue record generation module is used for executing closed-loop processing of the corresponding business on each line according to the reset business dialogue template, filing the current business dialogue to the corresponding business domain to obtain a business domain dialogue record set, and ensuring that each outbound business semantic link is kept independent.

The invention relates to the technical field of household system abnormity traceability, in particular to an intelligent household system abnormity traceability method and device based on cross-layer association reasoning, and the method comprises the steps: constructing an application-level and network-level behavior dependence graph; executing a cross-layer graph alignment strategy on the two types of graphs, establishing a cross-layer consistency semantic link, marking verification result features, and generating a feature enhanced network and application behavior graph; converting the network and application behavior maps into a structured fact library by utilizing fact codes, and deducing a cross-layer abnormal propagation path and a potential root cause; and converting into newly added semantic nodes and relation edges in the application behavior map, and finally generating a global traceability map fusing double-layer semantics and completely describing system behaviors and an abnormal propagation chain. Therefore, the problems that in the related technology, due to the fact that a cross-layer causal relationship is missing, a serious blind area exists in the coverage range of anomaly detection, misjudgment of a problem source can be caused, and positioning of a system fault is affected are solved.