51 results about "Knowledge conversion" patented technology

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.

Embodiments are directed toward a blockchain system including a global state and an assemblage of blocks, each block representing a collection of state transformation records, each state transformation record describing a state transformation performed on the global state, where preceding blocks referenced by any given block contain state transformation records describing state transformations performed on the global state prior to the evaluation of the given block, and where at least one state transformation record is a zero-knowledge transformation record encoding a zero-knowledge state transformation description, which zero-knowledge transformation record comprises at least the following elements: paths identifying the locations of the elements of a discrete data subset, a revised data subset, and a transition proof implemented as a non-interactive zero-knowledge proof, which transition proof proves that the transition from the discrete data subset to the revised data subset follows the established rules of the blockchain system.

The invention relates to the technical field related to medical systems, in particular to a medical teaching and research integrated system based on a common bottom layer, and aims to break through information islands among medical treatment, teaching and scientific research and realize closed-loop linkage and dynamic knowledge conversion of a whole process. The system comprises a diagnosis and treatment decision subsystem, a scientific research queue management subsystem and a teaching subsystem, integrates data through a three-in-one data engine, and realizes information sharing and cooperative work. The diagnosis and treatment decision subsystem provides diagnosis and treatment support for severe diseases based on an expert thinking chain. And the scientific research queue management subsystem is responsible for screening cases, sorting data, implementing quality control and statistical analysis, and adjusting an expert thinking chain according to scientific research results. And the teaching subsystem carries out teaching by using a standardized case presentation mode. Through a feedback closed loop of clinical discovery, scientific research verification, knowledge updating and teaching feedback, rapid propagation and application of medical knowledge are promoted, and the medical service quality and the medical education level are improved.

Embodiments are directed toward a blockchain system including a global state and an assemblage of blocks, each block representing a collection of state transformation records, each state transformation record describing a state transformation performed on the global state, where preceding blocks referenced by any given block contain state transformation records describing state transformations performed on the global state prior to the evaluation of the given block, and where at least one state transformation record is a zero-knowledge transformation record encoding a zero-knowledge state transformation description, which zero-knowledge transformation record comprises at least the following elements: paths identifying the locations of the elements of a discrete data subset, a revised data subset, and a transition proof implemented as a non-interactive zero-knowledge proof, which transition proof proves that the transition from the discrete data subset to the revised data subset follows the established rules of the blockchain system.

Computer-implemented methods for recommendation prioritization for a container orchestration system. Aspects include receiving a set of recommendations for a cluster of a container orchestration system. Aspects also include selecting an optimal recommendation from the set of recommendations using a scored knowledge transform graph. Aspects further include generating a confidence score for the cluster based on the optimal recommendation. Aspects also include determining a category of a readiness assessment model for the cluster using the confidence score. Aspects further include modifying a computer resource of the cluster based on the category of the readiness assessment model.

The invention relates to the technical field of employee training, and discloses an artificial intelligence-based employee training assessment method, device and equipment, and the method comprises the steps: obtaining the basic data, work behavior data and historical training data of an employee, carrying out the fusion of all data, and constructing a multi-dimensional employee portrait; based on the performance goals, determining training goals of the employees, and generating personalized training contents; a knowledge system is constructed according to personalized training content, real-time knowledge recommendation is performed in actual work, and a practice result is recorded; generating personalized assessment content for intelligent assessment to obtain an assessment result; and evaluating the employee training effect according to the practice result and the assessment result. The employee portrait is constructed through multi-source data fusion, accurate mapping from performance goals to personalized training is achieved, the matching degree of training content and employee ability short boards is improved, the knowledge conversion efficiency is enhanced, closed-loop management is formed in combination with real-time knowledge recommendation and scene practice, and the talent training efficiency is improved.

The invention discloses a heterogeneous AI engine interaction system and method based on a cognitive firewall, and relates to the technical field of artificial intelligence, and the system comprises a cognitive isolation channel which is used for building an isolation communication path between an RAG system and an Agent system, and generating standardized knowledge; the three-layer filtering architecture is used for performing grammar layer verification, semantic layer understanding and pragmatic layer evaluation on the standardized knowledge in sequence and outputting safety knowledge passing verification; the bidirectional value alignment module is used for mapping the safety knowledge to a multi-dimensional value space through a forward value alignment mechanism to calculate value conformity, and generating value alignment knowledge according to the value conformity; and the knowledge-instruction conversion engine is used for converting the value alignment knowledge into a structured instruction which can be executed by the Agent system. The method can actively identify, intercept and avoid potential dangerous, misleading, prejudice or ethical inconformity knowledge applications, and greatly reduce negative effects caused by wrong use of knowledge.

The invention relates to the technical field of image data processing, in particular to an oral cavity knowledge graph construction method and device, equipment and a storage medium. Preprocessing the oral CT original sectional image data based on a preset adaptive threshold method and a preset rigid registration algorithm to obtain standard image data; performing feature extraction on the standard image data based on a preset deep learning model and a preset principal component analysis method to obtain a structured feature vector; constructing the structured feature vector, the structured data and the unstructured data according to a preset entity linking method, a preset inference rule and a preset graph neural network model to obtain an optimized knowledge graph; an optimized knowledge graph is constructed through CT image preprocessing, deep learning feature extraction and multi-source data fusion, an image-to-structured knowledge conversion link is realized, and a comprehensive, reliable and convenient knowledge data support is provided for precise diagnosis and treatment of oral diseases.

The invention discloses a tunnel science popularization element universe system, and relates to a traffic informatization and science popularization digital information technology. A tunnel development display module, a panoramic exhibition hall display module, an activity synchronization module, a rescue channel module, a vehicle inspection module, an unmanned aerial vehicle inspection module, a robot fire extinguishing module, a robot dog inspection module, a token interaction module, a disease evolution module, a fire escape module and a visual induction experience module are integrated. Original abstract tunnel knowledge is converted into visual and vivid virtual experience, in addition, complex tunnel knowledge is dispersed to all the modules, a large amount of complex tunnel knowledge can be presented to a user in an organized mode, and the user can conveniently select tunnel science popularization knowledge needing to be known according to actual needs; and meanwhile, the user can interact with the virtual environment through the interaction equipment corresponding to each module, so that the interestingness of learning is improved while the user knows tunnel knowledge.

Data flows and data processing modules are provided to fulfill the implementation of: contextualized data collection, scalable analytics, and machine learning operations quality assurance. These modules may be implemented as standalone components, and / or bundled as a group of coordinated microservices. The conversion of raw domain expertise and knowledge into data processors, analytics, and visualization modules for industry oriented Industrial Internet of Things (IIoT) solutions is streamlined.

The invention belongs to the technical field of network security, and discloses a construction method of a network security knowledge graph based on SecureBERTPlus, and the method comprises the steps: achieving the knowledge conversion of threat intelligence through the cooperative processing of a data preprocessing module, a semantic information extraction module and a knowledge graph construction and optimization module; and a complete knowledge engineering link from original data to decision support is formed. According to the construction method adopted by the invention, synchronous optimization of entity recognition and relation extraction is realized through a joint extraction framework; semantic features of terms in the network security field are accurately captured through a field-adaptive SecureBERTPlus pre-training model, time sequence dependence in an attack chain is modeled by using a BiLSTM layer, a cross-sentence key relationship is focused by means of an attention mechanism, and ATTamp is injected based on a CRF layer; a CK rule constrains a decoding process, and a hierarchical clustering HAC based on context embedding is particularly introduced to realize entity dynamic disambiguation and semantic fusion.

The invention provides a resident health suggestion recommendation method, system and device based on a chronic disease knowledge graph and a large model and a medium, and belongs to the technical field of health suggestion recommendation. The method comprises the steps that resident health description of a user is responded, and chronic disease health keywords are extracted through a Lma model; coding chronic disease health keywords based on a Bert model, then aligning the chronic disease health keywords with entity words in the chronic disease knowledge graph, and extracting chronic disease related knowledge from the chronic disease knowledge graph; the chronic disease related knowledge is converted into a natural description language, and chronic disease related knowledge description is obtained; the chronic disease related knowledge description and the resident health description are assembled and then input into a Lama model for analysis, and the chronic disease with the maximum risk is obtained; and searching health suggestion knowledge of the highest-risk chronic disease from the chronic disease knowledge graph, and returning the health suggestion knowledge to the user. According to the method, key information is extracted from natural language description of residents by combining the LLaMA model and the Bert model, and accurate health suggestions are provided based on the knowledge graph.

The invention relates to a program design aided teaching agent system based on retrieval enhancement generation, and aims to solve the problems of knowledge fragmentation, low retrieval precision, deviation of generated content from teaching materials, insufficient teaching guidance and the like of a traditional teaching system. A knowledge unit construction technology combining hierarchical semantic segmentation and position coding, a precise matching mechanism combining two-stage hybrid retrieval and cross encoder reordering, a structured problem-document splicing fidelity technology and a multi-round progressive multi-round guide strategy based on a CDIO concept are adopted. Knowledge unit division with complete semantics is realized through an HAC-BERT + KMeans algorithm (the accuracy is improved by 28%), BM25 and dense retrieval weighted fusion (the weight is 0.4: 0.6) is combined with cross encoder reordering (NDCG at10 reaches 0.735), textbook format information (the content reuse rate is 87%) is reserved through structured splicing, and finally a complete technical scheme from knowledge organization and accurate retrieval to progressive teaching guidance is formed. And the intellectualization level and the knowledge conversion efficiency of program design teaching are obviously improved.

The invention relates to the technical field of electrical digital data processing, in particular to a process standardization-oriented production knowledge automatic extraction and management system, which comprises a data acquisition and processing module for acquiring multi-modal time sequence data, performing preprocessing and feature extraction and generating a multi-modal feature sequence; the skill identification module is used for segmenting a skill primitive sequence and acquiring a key parameter vector of each skill primitive based on the multi-modal feature sequence, and identifying an operation intention based on the skill primitive sequence and the key parameter vector; the skill modeling module is used for constructing a common skill model and a personalized skill model of the process operation process based on the skill primitive sequence; the model evaluation module is used for defining multi-dimensional evaluation indexes and evaluating operation effects corresponding to the generality and personalized skill models; and the knowledge management module is used for constructing a structured process knowledge base supporting dynamic updating and complex query. According to the method, a data-to-knowledge conversion system is established, and a scientific and efficient process standardization process is realized.

The application discloses a knowledge management method and system based on information intelligent retrieval, relates to the field of data processing, and comprises the following steps: importing historical input modal data of a permission user within a current period for retrieval; adopting a modal recognition algorithm to count the frequency of various types of modal data in the historical input modal data, and determining the dominant input modal of the permission user; constructing a knowledge conversion rule model according to the retrieval expression of the dominant input modal; connecting an initial knowledge base of an information intelligent retrieval system, adaptively converting the knowledge content in the initial knowledge base by using the knowledge conversion rule model, outputting a converted knowledge base, and searching the real-time input modal of the permission user in the converted knowledge base to return a search result. The application solves the technical problems of low search efficiency and insufficient matching degree of the existing knowledge management based on information intelligent retrieval, and achieves the technical effect of improving the search matching degree and efficiency.

The embodiment of the invention provides a model processing method and related equipment. The method comprises the steps of obtaining a to-be-trained first language model and a pre-training task constructed for the first language model; the training target of the pre-training task comprises the step of converting knowledge of a specified field into model parameters of a first language model and storing the model parameters; according to a serial execution sequence, sequentially executing N (an integer greater than 1) subtasks contained in the pre-training task to perform iterative pre-training on the first language model to obtain a second language model which is used for performing text classification processing based on knowledge in a specified field; the serial execution sequence is determined according to the task difficulty of each subtask; in the process of executing the ith sub-task to perform ith pre-training on the first language model, training data of the ith pre-training comprises a group of text sequences corresponding to the ith sub-task, and i is a positive integer smaller than or equal to N. The pre-training effect of the language model can be improved, and the classification performance of the language model is enhanced.

The invention discloses an implicit knowledge mining method and device and electronic equipment, and relates to the technical field of implicit knowledge mining, and the device comprises a perception data collection module, a preprocessing module, an implicit knowledge extraction module, an implicit knowledge conversion module and a knowledge structured display module; the perception data collection module collects different types of perception data, including video, audio and physiological data, and all the collected data are transmitted to the preprocessing module for subsequent processing and analysis; the preprocessing module cleans and processes the original data, and standardizes the time data by using a timestamp formatting tool; according to the method, the knowledge graph constructed by the implicit knowledge conversion module is integrated with the tagged data, and the Elasticsearch is combined to realize quick query and real-time push based on time, emotion and behavior dimensions, so that the data retrieval efficiency is greatly improved, the user is helped to timely obtain the required implicit knowledge, and the decision process and the user experience are optimized.

The invention discloses a question processing method and a related device, and relates to the technical field of large language models.In the question processing process, knowledge related to a question in a knowledge graph is converted into knowledge chain embedded representation, the knowledge chain embedded representation can reveal the incidence relation of all the knowledge, and the knowledge chain embedded representation can be used for representing the incidence relation of all the knowledge; the knowledge chain embedded representation is provided for the large language model, the large language model can refer to the incidence relation between the knowledge in the reasoning process to generate a reply to the problem, and therefore the representation of the knowledge in reasoning can be improved, and then the reasoning effect can be improved.

The utility model discloses a simulated fuel oil filling device, which comprises a rear oil tank, the front oil tank is arranged on the side wall of the rear oil tank; the small oil pool is arranged on the bottom surface of the rear oil tank; the rear oil tank and the front oil tank are communicated with each other through the flow guide pipe; the first valve is used for mounting a safety valve; the second valve is used for being connected with an excretion gun faucet; the third valve is used for being connected with a filling gun faucet; and the transmission pipeline is used for communicating the small oil pool with the rear oil tank. According to the utility model, through the arrangement of the oil tank and the valve, liquid water is utilized to simulate the fuel oil filling process, and teaching personnel can better teach the fuel oil filling principle for students by combining the whole simulation device, so that the students can more efficiently and smoothly master fuel oil filling, the knowledge conversion is more intuitively presented, and potential safety hazards are eliminated.

Train safety risk identification in a safety supervision scene is a core for constructing driving safety risk knowledge engineering and knowledge-driven railway intelligent safety risk management. How to automatically acquire train safety risk knowledge from massive unstructured data is a bottleneck of quick conversion from data to knowledge in a safety supervision scene. In order to solve the problem, firstly, a train safety risk knowledge ontology structure is formalized, and a complete knowledge system is constructed; the method comprises the following steps of: firstly, extracting a domain-adaptive knowledge extraction (DAKA), and then, proposing a domain-adaptive knowledge extraction (DAKA) method. According to the method, a double-flow self-attention mechanism is used, universal semantic embedding of safety supervision check text data is learned, then field features are continuously learned in a large-scale railway field text, and the characterization capacity of railway field semantics is improved; on the basis of understanding semantic features of domain data, a segmented convolutional neural network is used to deeply mine train safety risk features in a fine-grained manner, and finally, a train safety risk knowledge automatic acquisition task is realized.

The application discloses a knowledge acquisition method based on neural symbol hybrid retrieval enhancement, comprising the following steps: acquiring an event corpus, a semantic graph, and a symbolized knowledge base related to an event extraction task; mapping each corpus text in the event corpus and an event type in the symbolized knowledge base according to the semantic graph by a weak supervision mode to construct a retrieval knowledge base; when event extraction is performed on a target text, retrieving a plurality of candidate corpus texts related to the target text from the retrieval knowledge base, retrieving an optimal event type for the target text according to a mapping relationship between the candidate corpus texts and the event type, and extracting an event subgraph corresponding to the optimal event type and an optimal corpus text as prompt knowledge; combining the prompt knowledge converted into a prompt text with the target text as an input text, performing knowledge acquisition on the input text by using a language model, and obtaining an event extraction result based on the target text. The method enhances the accuracy and efficiency of knowledge acquisition.

The invention provides an LLM-based human resource intelligent training method and system, and the method comprises the steps: obtaining historical employee training information, and carrying out the preliminary processing, so as to train a first LLM training model; obtaining enterprise training demand information, outputting a training knowledge base according to the first LLM training model, training employees according to the training knowledge base, and obtaining a first training result; calculating a training matching rate according to a training result, and performing specific fine adjustment on the first LLM training model to obtain an enterprise adjustment model; calculating a staff knowledge conversion rate, obtaining staff knowledge short board information, setting a training vector according to the staff knowledge short board information, performing secondary specific fine adjustment on the first LLM training model through the training vector, and obtaining a staff adjustment model. Historical data utilization and model training and fine tuning are carried out to finally meet the double requirements of enterprises and employees.

This invention discloses a structured knowledge query system based on a large language model, comprising a knowledge representation module, a function generation module, and a query module. The knowledge representation module converts structured knowledge into a condition graph, which includes a set of nodes and a set of conditions. The function generation module generates a query function based on the query prompts and query questions using the large language model. The query module performs knowledge queries based on the function in the condition graph to obtain query results. This system combines the high accuracy of structured query languages ​​with the good generalization ability of LLMs, achieving accurate querying of various types of structured knowledge without data leakage.

This invention discloses a project compliance audit method and system based on collaborative reasoning using an expert knowledge base and a large model. First, a large model is used to analyze a collection of laws and regulations and audit practice documents, automatically extracting compliance audit-related knowledge. Then, this compliance audit-related knowledge is used to construct an expert knowledge base comprising a legal and regulatory knowledge base, an audit case library, an audit experience library, an audit practice library, and a compliance audit rule library. This invention achieves the function of automatically constructing and dynamically expanding the expert knowledge base by extracting legal and regulatory knowledge and audit knowledge using a large model. Furthermore, the reliability of the expert knowledge base and factual data is ensured through a confidence level mechanism and human-machine collaborative verification. Simultaneously, the use of audit evidence traceability chain information enables full-link traceability from audit conclusions to original evidence, greatly improving knowledge conversion efficiency and enhancing the interpretability and reliability of compliance audits, making it suitable for widespread promotion and use.

The application discloses a retrieval enhancement generation method and device, equipment, a storage medium and a program product. The method comprises the following steps: determining a mixed embedding model based on a main embedding model and a first knowledge base; the mixed embedding model comprises the main embedding model and at least one sub-embedding model; converting knowledge in the first knowledge base into vector representation of a corresponding language through the mixed embedding model to obtain a second knowledge base; each piece of knowledge in the first knowledge base is marked with a corresponding language label; retrieving the second knowledge base based on a retrieval question through the mixed embedding model to obtain a retrieval result set; the retrieval result set comprises retrieval results corresponding to multiple languages.

The application provides a model compression method and a computer readable storage medium. The method comprises the following steps: obtaining a sample image and label information, processing the sample image by using a teacher network to obtain a teacher intermediate layer feature map and a teacher prediction result; processing the sample image by using a student network to obtain a student intermediate layer feature map and a student prediction result; inputting the label information and the teacher intermediate layer feature map into a knowledge conversion network to obtain a target feature map and a background feature map, extracting key information from the target feature map and the background feature map respectively to obtain a target key feature map and a background key feature map; iteratively training the knowledge conversion network and the student network according to a total loss function; repeating the steps of extracting the intermediate layer feature, the step of extracting the key feature and the step of knowledge migration until an iteration stopping condition is reached, taking the student network as a compressed model to complete the knowledge migration from the teacher network to the student network and the compression of the network model.

The invention relates to the technical field of artificial intelligence and medical fusion, in particular to a multi-modal AI agent platform which comprises a knowledge conversion engine, a medical knowledge reasoning model and a medical intelligent application model. Wherein the knowledge conversion engine is composed of an Agent, an Orchestrate, a Routine and a Function call; the Agent is used for constructing a medical service field, performing man-machine interaction, performing a multi-Agent collaborative management system and integrating medical tasks; orchestrate is used for completing medical service decoupling fragmentation, arrangement innovation and multi-service collaboration, and promoting medical service fusion innovation and cross-domain rearrangement. According to the invention, the pain points of system splitting, difficulty in innovation, high cost, difficulty in fusion and the like in the medical field can be effectively solved, intelligent upgrading of medical application, flexible rearrangement of field application and cross-field deep fusion are realized, and multiple values of remarkable reduction of medical cost, efficient innovation of medical knowledge, optimization of medical services and innovation of medical business modes are achieved.

The present invention discloses a network collaborative knowledge management method and system, which includes: classifying and refining the device data to be processed, obtaining multiple knowledge classifications, and the multiple knowledge classifications contain multiple fine-grained knowledge; wherein the multiple fine-grained knowledge contain unstructured knowledge, traversing the multiple fine-grained knowledge, and using hierarchical classification rules to convert the unstructured knowledge into structured knowledge; using a tree structure to construct a knowledge graph from the multiple knowledge classifications and the multiple fine-grained knowledge, wherein the multiple knowledge classifications serve as the root nodes of the knowledge graph, and the multiple fine-grained knowledge serve as the leaf nodes of the knowledge graph. The present invention greatly improves the efficiency of retrieval and reuse of device data by acquiring fine-grained knowledge and converting unstructured knowledge into structured knowledge to form a knowledge graph.

The embodiments of the present application disclose a method, apparatus, medium and equipment for constructing a failure case knowledge management system, which relates to the field of data management technology, including: performing knowledge conversion on text data to obtain failure keywords; performing embedding construction on parameter signal data to obtain embedded data; generating rules based on the embedded data using a large language model to obtain failure rules; and constructing a failure knowledge network based on failure keywords and failure rules to realize the construction of a knowledge management system. Since the present application introduces embedding technology in the conversion, it can realize the non-aligned knowledge conversion of multimodal data of traditional text data and parameter signal data, and the embedding technology includes three embedding means to convert important time series information such as the relationship and time series of multi-dimensional monitoring parameters into knowledge, while retaining the time series characteristics, improving the learning ability of the knowledge process for the time relationship in the failure process, and improving the management effect of the knowledge management system of failure cases.