13883 results about "Linguistic model" patented technology

Provided in the present application is an intelligent question answering method based on a collaboration between a large language model and a knowledge graph, relating to the technical fields of artificial intelligence and natural language processing, the method comprising: decomposing a complex question into a plurality of simple questions, and analyzing the degree of association between the simple questions and a basic function so as to form a multi-hop reasoning path; automatically extracting structured information from the simple questions on the basis of a multi-task learning framework of a large model, so as to construct a knowledge graph; and constructing a cumulative reasoning learning framework on the basis of a logic reasoning large model, and performing iterative verification on a process result formed by the knowledge graph on the basis of the multi-hop reasoning path, so as to correct the reasoning path until a correct answer is inferred.

The invention discloses a power plant operation and maintenance knowledge intelligent query method based on a large language model and an RAG technology. The method comprises the following steps: constructing a power plant operation and maintenance knowledge vector library covering structured, semi-structured and unstructured data; receiving a natural language question of a user, inputting an improved instruction to align a preprocessor, and generating a question semantic vector and an intention tag; relevant knowledge fragments are retrieved and sorted through a semantic matching retriever in combination with the intention labels; constructing a large language model cue word structure based on the retrieval result and the original question, generating candidate answers and recording a reference path; and finally, performing term specification and consistency verification according to the expert rule base, and outputting a structured and traceable final answer. According to the invention, the improved RAG technology is fused to realize intelligent query of the operation and maintenance knowledge of the power plant.

In a described embodiment, a system for information processing is provided including a data acquisition module configured to receive feedback corresponding to one or more outputs generated by a language model. The system further includes a cognitive reasoning module configured to evaluate the reasoning process of the language model, emulate cognitive functions including metacognitive processes, and generate an assessment based on an analysis of the received feedback, wherein the assessment includes classifying the one or more outputs into components, assigning quality scores for each component, and identifying an improvement corresponding to the one or more outputs. Additionally, the system includes a process adjustment module coupled to the cognitive reasoning module for adjusting the reasoning process of the language model based on the assessment is provided. A refinement module coupled to the process adjustment module is provided for iteratively refining the reasoning process based on subsequent updates to the generated assessment until a performance threshold is met.

The invention relates to the technical field of financial data processing, in particular to a multi-source heterogeneous fund data processing method and system, and the method comprises the steps: recording source information through building a data source registry, and marking a unique identifier for data; performing differential analysis on the fund data in different formats to generate standardized column type storage data; traversing column type storage data to extract statistical characteristics, performing field classification through metadata analysis and financial dictionary matching, analyzing business connotations of fields difficult to classify in combination with a localized large language model, and mapping the business connotations to a header fusion knowledge graph; generating a mapping rule from the source field to the enterprise-level data model by applying a rule engine template on the basis of field classification and semantic recognition results; converting the data structure according to the mapping rule and executing standardization processing; the quality is further optimized through data cleaning; and finally, the data are verified, standardized fund data supporting data traceability are output, and the strict requirements of financial supervision application are met.

Disclosed in the present application are a multi-modal data processing method and apparatus, an electronic device, and a storage medium. The method comprises: acquiring a reference image and a reference text; extracting a reference visual feature of the reference image; by means of a multi-modal large language model, determining an embedding of the reference text, an embedding of a start mark of the reference visual feature, an embedding of the reference visual feature, and an embedding of an end mark of the reference visual feature; on the basis of the multi-modal large language model, splicing the embedding of the reference text, the embedding of the start mark, the embedding of the reference visual feature, and the embedding of the end mark into a target embedding sequence, performing attention processing on the basis of the embedding of the start mark, the embedding of the end mark, and an embedding selected by a sliding window in the target embedding sequence, and outputting a predicted sequence; and generating a predicted image and a predicted text on the basis of the predicted sequence.

The invention relates to a large language model construction method and system fused with spatial semantic understanding, and the method comprises the steps: obtaining a multi-source heterogeneous corpus, and extracting an entity, an attribute and a business rule; extracting a semantic feature vector set based on the multi-source heterogeneous corpus, and constructing an entity relationship network and an enhanced knowledge graph; generating an enhanced training sample, and training the general large language model to obtain a primary large language model; generating a verification sample set and performing verification; identifying a specific weakness pattern, and generating a corresponding confrontation sample and a knowledge enhancement sample; training the primary large language model to obtain an optimized large language model; in conclusion, the enhanced knowledge graph fusing the spatial semantic features and the business rules is constructed, and the gradient training samples are generated based on the graph to perform multi-stage model training and optimization, so that the method has the effects of improving the internalized understanding ability of the model for the spatial semantics and the business rules and enhancing the reliability of multi-step spatial reasoning.

The invention discloses a real-time computing system resource coordination and decision engine system based on a large language model. According to the system, a large language model is innovatively used as a central strategic decision engine, and a'decision-coordination-execution 'three-layer architecture is constructed. According to the system, macroscopic strategy generation and microscopic real-time control are decoupled by introducing a hierarchical decision-making mechanism (a strategic layer, a tactical layer and an execution layer), so that the core contradiction between LLM high reasoning delay and the microsecond / millisecond-level real-time requirement of the system is effectively solved, and the method is suitable for local computing equipment and a cloud data center. The system comprises a predictive strategy preloading system, and transient response can be achieved. Meanwhile, the system adopts an asynchronous event-driven decision-making mechanism for continuous intelligent optimization. According to the method, the top-down, semantic understanding-based and global collaborative intelligent management of the computing resources is realized, and the resource utilization efficiency, the system automation degree and the overall energy efficiency in a complex and dynamic computing environment are remarkably improved.

The invention belongs to the field of network security, and particularly discloses a network defense agent system based on a large language model. Through the design of the sensing layer, the decision analysis layer and the action execution layer, comprehensive protection of network threats is realized. The sensing layer is responsible for collecting original information from multiple channels and converting the original information into standardized data; the decision analysis layer performs modeling and threat reasoning on attack behaviors, evaluates a risk level and predicts subsequent actions; and the action execution layer specifically executes defense operation according to the defense strategy scheme output by the decision analysis layer. In addition, the application also constructs a data set oriented to attack and defense confrontation, records a complete attack sequence, defense response and effect evaluation thereof, and provides a reliable basis for continuous learning of defense agents. Experimental results show that the framework provided by the invention is superior to the traditional method in the aspects of attack detection accuracy, attack chain identification and defense strategy generation, and has stronger adaptability and real-time response capability.

The invention discloses a structured decision-making method based on multi-agent collaborative decision-making and reinforcement learning, and relates to the technical field of natural language processing, knowledge engineering and agent collaboration, and the method comprises the steps: receiving an original rule document, analyzing the document type, complexity and constraint conditions, and defining a task target and a success standard; and according to the task target, matching and scheduling the intelligent agent from the registered intelligent agent library, and further analyzing the capacity configuration of the intelligent agent for standby. Through the multi-agent cooperation and reinforcement learning technology, full-process automation of rule documents from input to structured analysis is realized, document types, complexity evaluation and constraint condition analysis can be automatically identified, and a clear task target and a success standard are generated; and the large language model generates a structured workflow according to task requirements and agent capabilities, so that the performability is ensured through logic verification, manual intervention is greatly reduced, and the processing efficiency and the system intelligence degree are improved.

Systems, methods, and devices that relate to routing requests to large language models (LLMs) are disclosed. In one example aspect, the system receives session-specific data elements in response to a request to generate an output using LLMs. The system determines a hierarchy of operational constraints including privacy protocols and performance requirements. Weights for a multi-variable optimization are dynamically updated using the session-specific data elements. The system executes the multi-variable optimization across candidate LLMs that satisfy privacy constraints and optimize performance constraints. Based on the optimization, at least one candidate LLM is selected and the request is routed to it. In response to performance feedback, the system automatically selects a different LLM to improve one constraint, resulting in degradation of another constraint.

The invention provides an enterprise process intelligent analysis system based on a large language model. The enterprise process intelligent analysis system comprises a master control scheduling module, a data preprocessing module, a hierarchical analysis module, an insight extraction module, a report generation module and a knowledge retrieval module. The master control scheduling module generates a scheduling plan based on chain thinking reasoning, and dynamically calls each module; the data preprocessing module carries out cleaning and structured conversion on the enterprise event logs and outputs standardized JSON (JavaScript Object Notation) data; the knowledge retrieval module is combined with an RAG technology and a vector database to provide context support for a large language model; the hierarchical analysis module drives a model to execute process discovery and bottleneck identification through a structured cue word template; the insight extraction module converts an analysis result into a commercial insight text containing reasons, influences and suggestions, and has a self-repairing mechanism to guarantee consistency; and the report generation module automatically generates an image-text report. The system can improve the efficiency and accuracy of process analysis.

The invention relates to the technical field of database operation and maintenance, in particular to a database operation and maintenance decision-making method and device based on a large language model.The method comprises the steps that real-time operation indexes and historical log data of a database are collected, and multi-dimensional time sequence features are extracted; and analyzing the characteristics by using a large language model, judging whether a performance bottleneck or an abnormal risk exists or not, and determining the relevance between the performance bottleneck or the abnormal risk and a database performance problem. If the risk exists, acquiring a system load sudden change trend and a resource request rate, quantifying a system pressure degree, and inputting a performance optimization model to generate an adaptive tuning strategy; calculating statement complexity through an SQL audit log, and selecting an index adjustment strategy in combination with an index optimization model; evaluating the influence of the tuning strategy and the index strategy on the throughput, the response delay and the resource utilization rate, and determining a target high-performance strategy; according to the scheme, accurate optimization and efficient operation and maintenance of database performance are realized through intelligent analysis and dynamic adjustment and optimization.

The invention relates to the technical field of natural language processing and intelligent content generation, and discloses an automatic generation method and system for a 3D scene interaction script based on a large language model.The method comprises the steps that input character description is obtained, role and event information is extracted through semantic analysis, and a multi-role interaction graph is constructed to obtain a preliminary script branch; calculating the matching degree of the node and the world view, and optimizing the path and the node attribute when the matching degree is low; logic verification points are extracted to verify the continuity of the plot; embedding role emotion to generate initial plot segments; and optimizing content connection and dialogue rhythm, and fusing plot promotion elements to output a final script. The method realizes automation and high quality of script generation, ensures plot coherence, logic compliance and natural emotion, and improves the immersion experience of the user.

The invention provides an intelligent data query method based on a natural language, and relates to the technical field of intelligent data processing and natural language interaction.The intelligent data query method comprises the steps that firstly, enterprise original data is subjected to standard treatment, and a standardized theme database and a data directory and index definition document are constructed; key semantics are extracted based on unstructured knowledge, and a domain knowledge vector library is fused and constructed by combining document text fragments and vector representation of a mapping relation between historical questions of a user and an SQL (Structured Query Language). And after receiving a natural language question of a user, calling a large language model to identify a task type, and distinguishing knowledge questions and answers, data query and complex analysis. Executing corresponding operations according to different types: directly retrieving a vector library by knowledge questions and answers to generate answers; extracting keywords in data query and generating a query request in combination with context; and in the complex analysis, predefined workflow is judged and executed or intelligent agent processing is called, and query or analysis requirements are output.

The invention discloses a low-code development automatic generation method based on a large language model, which comprises the following steps: collecting natural language description information input by a user, and preprocessing; inputting the standardized demand corpus set into a large language model, and executing semantic understanding and context modeling; matching the low-code component library based on the structured semantic representation to generate component assembly description information; generating and verifying an engineering skeleton according to the assembly description information, and outputting an executable low-code application initial version; running the executable low-code application initial version, and monitoring and analyzing execution difference to generate an increment adjustment instruction; and inputting the increment adjustment instruction into the large language model, performing reconstruction and adaptive optimization, and outputting an executable low-code application final version. According to the method, large language model semantic understanding and adaptive optimization technologies are fused, automatic generation and continuous optimization of low-code applications are realized, and the method has the advantages of intelligence, high precision and engineering reliability.

The invention belongs to the field of question and answer generation, and provides an interactive retrieval enhancement question and answer generation method and system based on a knowledge graph, and the method comprises the steps: carrying out the document partitioning based on an original document set, generating a global block set, carrying out the entity extraction of each text block in the global block set, and obtaining an entity set; performing relation extraction on entity subsets in each text block in the entity set to obtain a global relation set; generating a plurality of sub-knowledge maps based on the global block set, the entity set and the global relationship set, and performing entity fusion and relationship fusion on the sub-knowledge maps to obtain a knowledge map; performing keyword extraction and semantic embedding on the original problem to obtain a dense vector, performing semantic embedding based on the knowledge graph to obtain an embedded vector, and generating a candidate entity set according to the dense vector and the embedded vector; and based on the candidate entity set, utilizing a large language model calling tool to carry out extended search to generate a candidate information set, and utilizing a large language model to obtain an answer to the original question based on the candidate information set.

The invention discloses a software multi-agent collaboration method and system based on a large language model, and the method comprises the steps: receiving natural language task description submitted by a user at the same time, carrying out the semantic understanding and intention recognition through a pre-trained large language model center, and generating a structured task element set; based on the structured task element set, the large language model center generates a task dependency graph through multiple rounds of reasoning, and the task dependency graph comprises a plurality of atomic subtasks, logic relations among the tasks and data flow constraints; according to a topological structure and resource demand characteristics of a task dependency graph, a double-layer graph attention network is adopted to dynamically match a professional agent with specific domain capability, and a distributed collaborative network is formed. Through the dynamic graph network scheduling and cross-domain semantic alignment mechanism, the problems that the multi-agent dynamic collaborative adaptation capability is insufficient and cross-domain semantic fusion is difficult are solved.

The invention belongs to the field of artificial intelligence, particularly relates to a large model illusion suppression method, system and equipment based on a dynamic knowledge base and multi-modal consistency constraint, and aims at solving the problem that factual illusion is likely to occur when an existing large language model generates content. The method comprises the steps that a knowledge base of multi-source heterogeneous data is constructed and dynamically maintained, and a dynamic credibility weight fusing data source authority, knowledge timeliness and multi-modal consistency is calculated for each piece of knowledge in the knowledge base; when the content is generated by the model, high-credibility related knowledge is retrieved from the knowledge base according to the current context; in the decoding stage of the model, a constraint loss item is designed, and the generation probability is adjusted in real time by calculating the similarity between the currently generated content and the retrieval knowledge in the feature space. According to the method, the multi-modal knowledge base for dynamic credibility evaluation is introduced, and the real-time consistency constraint is applied in the generation and decoding link, so that the accuracy and the reliability of the generated content are remarkably improved.

According to the artificial intelligence-based data analysis system and method provided by the invention, a set of novel data analysis system architecture is designed, and key technical components such as natural language processing, a large language model, database query optimization, multi-modal visualization and a semantic knowledge graph are fused; data transmission between modules is realized through unified intermediate data objects such as a structured query intention, an analytic tree and a structured query language template object, and asynchronous collaboration is realized through event driving and a message queue mechanism, so that a non-professional user can input an analysis request through a natural language; semantic analysis, structured query language query statement generation, data query and result visualization presentation are automatically completed, the method can be widely applied to data analysis scenes in the industries of government affairs, traffic, finance, education and the like, the data use efficiency is improved, the technical threshold is reduced, and the digital decision-making ability is enhanced.

The present invention relates to the related technical field of natural language inference. Disclosed is a large language model joint inference method based on a knowledge graph-enhanced chain-of-thought prompt, comprising: constructing a local knowledge subgraph; decomposing an original question text into S sub-question texts and concatenating the original question text and the S sub-question texts; inputting a concatenated question text into a graph inference model to obtain a weighted entity distribution; from the weighted entity distribution, extracting first G answer entities having the highest confidence level, using an inference transition matrix to trace an inference process of each answer entity, and generating an inference path from a question entity to the corresponding answer entity; and using the inference path to assist a large language model in predicting an answer to the original question text. By means of the method, the large language model can quickly and accurately find an answer to a question text.

The invention discloses a laboratory quality management document intelligent generation method and system based on retrieval enhancement, and relates to the technical field related to data processing.The method comprises the steps that semantic coding is conducted on a preset standard text, and a vector knowledge base is constructed; retrieving associated standard terms according to the document theme, and extracting structured data from a laboratory business system; embedding the standard terms and the business data into a Prompt template, and calling a preset large language model to generate a text; and performing paragraph splicing and hierarchical control on the generated text, automatically checking compliance by utilizing term consistency of rule model fusion and a numerical value comparison algorithm, and outputting a quality management document. The technical problems that in the prior art, standard term retrieval and matching are not accurate, laboratory business data fusion is difficult, and consequently document compiling efficiency and quality are poor are solved, and the technical effects that minute-level automatic generation of laboratory quality management documents is achieved, and document compiling efficiency, quality and compliance are improved are achieved.

The invention discloses a public opinion risk assessment method and system based on multiple agents and a large language model, and belongs to the technical field of network information security. The method comprises the steps of obtaining multi-source public opinion data; reasoning in combination with a language model to obtain a multi-dimensional semantic vector, and clustering to form a plurality of topic clusters; the emotion opposition level, the credibility and the text quantity increment score of each topic cluster are generated based on an intelligent agent, a comprehensive risk value is obtained through weighted fusion, and high-risk topics are screened through a double-threshold retention mechanism; and generating a knowledge graph according to the high-risk topic, determining an associated entity, a propagation link and an intervention node, implementing an intervention strategy, and generating a public opinion intervention report. According to the method, the monitoring problem of multi-source heterogeneous public opinion data can be effectively solved, the accuracy and timeliness of risk assessment are improved, full-link automation from risk identification to accurate intervention is realized, and efficient support is provided for public opinion management and control of governments, enterprises and other mechanisms.

The invention relates to the technical field of natural language processing, in particular to an enterprise number asking system and method based on combination of a large language model and an NL2SQL, and the method comprises the following steps: filling a prompt project template with a natural language query request and a structured metadata context, and generating a standard prompt; the standard prompt is input into the large language model, deep semantic analysis is carried out, and an SQL query draft is generated; self-evaluation is carried out on the questions, and when the questions are recognized, clarified questions are generated and returned to the user; receiving feedback of a user, and iteratively optimizing the SQL query draft based on the feedback; performing grammar verification and security verification on the final SQL query draft to generate an executable SQL statement; sending the executable SQL statement to a target enterprise database for execution; and the information is visually displayed to a user. The problem that the SQL generated in the prior art deviates from the real intention of a user and cannot meet the requirements of enterprise-level applications for accuracy and reliability can be solved.

The invention discloses a multi-modal data driven general report generation method and system based on a large model, and belongs to the technical field of intelligent report generation. Firstly, texts, images and sensor data related to a report theme are obtained and subjected to standardized preprocessing; analyzing the report generation instruction, and matching and querying a task modal mapping library matching modal configuration scheme according to a task demand; quantitatively evaluating the data quality of each modal, dynamically calculating the final decision weight of each modal in combination with the basic weight, and distributing the final decision weight to a corresponding processing path to form dominant, supplementary and reference data; inputting the dominant data and the supplementary data into a multi-modal model for analysis to obtain a preliminary conclusion with confidence score, performing consistency judgment, if no conflict exists, performing fusion to form a comprehensive conclusion, and if the conflict exists, combining a quality evaluation result and an arbitration rule to complete conflict judgment; and finally, inputting the comprehensive conclusion and the reference data into a large language model to generate a report text, and outputting a complete report after typesetting and proofreading.

The invention relates to a multi-mode large language model fine tuning method, system and device and a medium, and belongs to the technical field of artificial intelligence and computer vision crossing. The fine tuning method comprises the steps that an original business scene image is acquired and preprocessed, and a preprocessed image is obtained; performing bounding box coordinate labeling and semantic label definition on the entity target in the preprocessed image through a labeling tool, and outputting a structured labeling file; based on the preprocessed image and the structured annotation file, constructing a training sample set comprising multiple rounds of image-text dialogues; loading the pre-trained multi-modal large language model, configuring low-rank matrix decomposition parameters, and generating a fine tuning instruction set; and inputting the training sample set into a pre-trained multi-modal large language model, carrying out joint training operation based on the fine tuning instruction set, and outputting the fine-tuned multi-modal large language model. According to the method, the identification accuracy, the interaction capability and the system availability of the visual question-answering system in an actual application scene are improved.

The invention provides an end-side multi-modal large model accelerated reasoning method and system, and the method comprises the steps: carrying out the two-stage screening and rearrangement of visual tokens based on the CLS attention and text-to-visual attention in a visual encoder and pre-filling stage, and constructing a sparse attention and sparse key value cache; in a decoding stage, an important neuron set is judged according to activation gating or historical statistics, only a corresponding feedforward network weight is pulled and calculated, missed weights are loaded on demand through asynchronous I / O, and hot neurons are maintained in a high-speed memory to utilize model sparsity, so that video memory / memory occupancy and calculation overhead are remarkably reduced on an end side; throughput and time delay performance are improved. According to the method, the internal memory and computing resources required by reasoning of the multi-modal large language model are reduced from two dimensions by utilizing the endogenous sparsity of the end-side large language model in input and the model, so that a higher reasoning speed is achieved by utilizing fewer resources on the premise of keeping the size of the model unchanged, and the performance of the whole system is improved.

The invention discloses an intelligent agent automatic arrangement method and system based on a large language model, and relates to the technical field of artificial intelligence. The method comprises the following steps: decomposing a natural language instruction of a user into a structured subtask sequence by utilizing a first large language model; based on the agent portrait library, matching and allocating agents for each sub-task to generate an initial execution plan; task execution is scheduled and monitored in real time through an event-driven architecture; when abnormity is monitored, a self-adaptive adjustment mechanism is triggered, the affected plan part is re-planned, and an updating instruction is issued. According to the invention, efficient, flexible and robust multi-agent automatic arrangement is realized.

The invention relates to the technical field of artificial intelligence, can be applied to the fields of financial science and technology and medical health, and discloses a reasoning optimization method, device, equipment and medium for a multi-modal large language model.The method comprises the steps that an input long context sequence is obtained, and key value projection is conducted on the long context sequence to generate an initial key value cache; for each attention layer of the multi-modal large language model, calculating an attention matrix of the attention layer according to the vector dimension of the long context sequence and the initial key value cache; calculating a cross-modal attention entropy according to the attention matrix, and determining a cache size of an attention layer according to the cross-modal attention entropy; optimizing the initial key value cache based on a cumulative attention scoring mechanism and a window strategy to obtain a target key value cache; and reasoning the long context sequence according to the cache size and the target key value cache to generate a long context reasoning result. And the reasoning efficiency and the reasoning accuracy are improved.

Techniques for generating an executable API call for an LLM-generated request, where the executable API call is usable to cause a component to generate a potential response to a user input, are described. In some embodiments, the system receives a user input and uses a language model to generate a request for a component to provide a potential response to the user input. The system uses the request, an API description corresponding to the component, and other information not available to the language model during processing to generate an executable API call corresponding to the request. The system can execute the executable API calls (in a system-determined order or concurrently) to cause the corresponding components to generate potential responses to the user input.

The invention relates to a large-scale road network traffic control method based on a deep reinforcement learning large model, and belongs to the technical field of intelligent traffic control. The method comprises the following steps: sensing real-time multi-modal road network information including urban road intersections, highway entrance ramps and emergency lanes, and generating a space-time fusion representation vector representing a current traffic network state by fusing a space diagram construction method and a time sequence embedding method; the space-time fusion representation vector and historical state memory are spliced to serve as input, a backbone network of a pre-training large language model is used for state feature distillation so as to enhance state representation, and a traffic control decision is output through a strategy network with a layered action space; through cross-modal knowledge migration and a progressive course learning strategy, a training process of a deep reinforcement learning algorithm is guided and optimized so as to improve model training efficiency and generalization ability. According to the method, the generalization performance and the accuracy of the control strategy are improved while the real-time response speed is ensured.