362 results about "Triplestore" patented technology

The invention discloses a data knowledge-based method based on semantic fusion, and relates to the technical field of computer information processing.The method comprises the steps that a requirement set serves as input, knowledge requirement analysis, concept modeling, relation modeling and constraint declaration are completed, and a semantic model OB is constructed; taking the original data set Sraw as input, completing standardization processing and structure segmentation under the support of a semantic model OB, and forming an entity corpus and a semantic unit set; based on the semantic unit set, structured and unstructured triple extraction, semantic verification and graph loading are executed, and an initial knowledge graph is constructed; performing entity alignment, relationship merging, rule reasoning and versioning release on the initial knowledge graph to generate a graph; introducing a quality evaluation mechanism, and outputting an optimized atlas and an evaluation document; the KG opt deployment is online, and query packaging, visualization, service arrangement and incremental maintenance are completed. The method aims at solving the problems that multi-source heterogeneous data are not uniform in structure and inconsistent in semantics.

The invention discloses a multi-table combined natural language query SQL (structured query language) generation method, which comprises the following steps of: 1, constructing a database meta-knowledge graph, and establishing a triple storage comprising a table structure, a primary and foreign key relationship and business description for each data table; 2, receiving a natural language query request, and calculating the topic relevancy between query semantics and each database table through a pre-trained topic matching model; 3, dynamically constructing a view, and logically associating the database tables of which the theme relevancy exceeds a threshold value to form a temporary view; 4, generating a context enhancement prompt, and combining the temporary view structure, the field semantic description of the view and the view content sample to form a structured prompt; 5, inputting the natural language query and the structured prompt into the large language model to generate candidate SQL statements; and step 6, executing verification and iterative optimization on the candidate SQL statements, verifying logic correctness through SQL execution plan analysis and result sampling, and triggering and prompting a reconstruction mechanism when detection is abnormal.

According to the multi-modal automatic knowledge graph construction method based on the large language model, a multi-modal data stream is preprocessed, features are extracted, and the multi-modal data stream is mapped to a unified semantic space through a cross-modal alignment network after being processed through the large language model, a visual converter and a time sequence neural network. In the space, entities and categories are recognized based on a large language model, a triple is generated by combining a multi-modal feature judgment entity relationship, mapping fusion is performed through an ontology alignment algorithm driven by a graph neural network and a predefined domain ontology, finally knowledge is stored in a graph database, and dynamic updating is performed by means of incremental learning and online reasoning. Standardized APIs and visualization components are provided. According to the method, the construction efficiency and the automation degree of the knowledge graph are remarkably improved, the cross-modal information fusion and knowledge maintenance capability is enhanced, and the application requirements of intelligent retrieval, recommendation, decision support and the like are met.

The invention belongs to the technical field of equipment knowledge engineering and natural language processing, and discloses an LLM-based constraint and self-repairing uniformization equipment knowledge graph automatic construction method. The method comprises the following steps: firstly, acquiring equipment related document data through network collection, document arrangement and database query; then, utilizing an equipment domain ontology and constraints as preposed soft and hard constraints, driving LLM to generate semantic intermediate representation, and obtaining candidate triples through structured compiling; then, a self-repairing closed loop is formed through semantic unit testing, logic consistency detection and evidence binding verification, and triples which violate constraints and have conflicts or illusions are automatically recognized and repaired; and finally, entity standard identification, cross-document duplicate removal combination and conflict resolution are realized through cross-segment unification and incremental alignment. According to the method, the fragile path that the LLM directly generates the triple and blindly stores the triple is avoided, the illusion and inconsistency problems are effectively inhibited, and the correctness, interpretability and maintainability of the equipment knowledge graph are remarkably improved.

The invention discloses a multi-agent cooperation enhancement method, system and equipment based on a knowledge graph, and the method comprises the steps: obtaining original data in an external environment, carrying out the preprocessing and feature extraction of the original data, generating a knowledge triple, storing the knowledge triple in a local knowledge graph, and submitting the knowledge triple to a shared knowledge graph for knowledge updating; when a to-be-executed task is received, decomposing the to-be-executed task by utilizing the large language model and querying global knowledge in the shared knowledge graph and local knowledge in the local knowledge graph to obtain a plurality of sub-tasks; a bipartite graph minimum cost matching algorithm is adopted to match a plurality of sub-tasks with the capability and availability of each agent to generate a preliminary task allocation scheme, and a large language model is utilized to optimize the preliminary task allocation scheme to generate an optimal task allocation scheme; and sending each task allocation knowledge fragment in the optimal task allocation scheme to a corresponding agent for collaborative execution through a semantic communication protocol.

The invention discloses a large-model-driven automatic knowledge graph construction method based on a confidence feedback mechanism, and aims to improve the structural accuracy and semantic consistency in a structured triple generation process, and perform structural constraint guidance by using a few-sample prompt mechanism and a cross validation mechanism of a heterogeneous large model. And the control capability of the large language model on the triple format is enhanced, so that format offset and semantic redundancy in the generation process are reduced. And meanwhile, a multi-dimensional confidence evaluation system is constructed, model consensus judgment, semantic rationality analysis and knowledge consistency verification are fused, and refined quantification and screening of triple quality are realized. According to the method, a confidence backtracking feedback strategy is introduced, a generation-verification-optimization closed-loop process is constructed, the expression and correction capability of the system on a complex knowledge structure is enhanced, the dependence on an external API is effectively reduced, the consumption of computing resources is reduced, and the operation efficiency of the system and the feasibility of engineering deployment are remarkably improved.

The invention provides a tunnel risk reasoning method fusing a knowledge graph and a large language model, which comprises the following steps of: obtaining structured monitoring data and unstructured text data, adopting methods such as field standardization for the structured monitoring data to realize a unified format, adopting methods such as sentence segmentation and word segmentation for the unstructured text data to realize the unified format, and obtaining the structured monitoring data and the unstructured text data; the method comprises the following steps of: extracting entities from data by utilizing a model, extracting a relationship between the entities based on the entities, forming basic triads, forming a sub-graph by the basic triads, integrating to form a knowledge graph, generating a natural language, extracting the sub-graph related to the natural language from the knowledge graph, and converting the sub-graph into a sub-graph in a vector form by utilizing a graph embedding algorithm. The entities and the relation paths of the entities serve as explicit reasoning clues, the natural language, the sub-maps in the vector form and the explicit reasoning clues are input into a large language model, natural language output is generated, multi-source data information is integrated, and high-precision and interpretable tunnel risk early warning is output through the large language model.

The invention discloses a water conservancy knowledge graph intelligent question-answering system and method based on a large language model, and relates to the technical field of water conservancy information, and the method comprises the following steps: carrying out the fusion processing of multi-source water conservancy data in advance, constructing a triple knowledge graph containing a water conservancy field entity type and a relation system, and dynamic updating of the knowledge graph is realized through an incremental learning algorithm. According to the method, the defects of a traditional method in semantic understanding are effectively overcome, deep semantic association of professional query can be accurately captured, and answer deviation caused by keyword matching limitation is avoided. Meanwhile, a dynamic updating mechanism of the knowledge graph can integrate new knowledge such as new projects and industry standard updating in real time through an incremental learning algorithm and a time decay function, obsolete out-of-time information synchronously, ensure that a knowledge system of the system is synchronous with development of the water conservancy industry, and solve the problems that a traditional system is high in updating cost and long in period.

The invention discloses an intelligent construction and tracing method and device for an attack graph, and relates to the technical field of network security. The method comprises the steps of performing semantic analysis and entity relationship extraction on a multi-source heterogeneous security log according to a predefined structured security data model, and generating a standardized security entity relationship triple set; based on the set, taking an entity in an initial alarm as a starting point, and adopting an iterative closed loop driven by a large language model to dynamically construct an attack graph; and carrying out attack technique and tactics mapping and threat attribution based on the final map, and generating a response strategy of priority ranking. According to the method, automatic and high-precision source tracing and response of the attack chain are realized, and the problems that the prior art depends on static rules and semantic segmentation and lacks dynamic reasoning capability are effectively solved.

The invention discloses a knowledge graph completion method based on combined prediction of large and small models, which comprises the following steps: 1, constructing and preprocessing a knowledge graph completion reference data set, and training by adopting a RotatE model to obtain candidate entities generated by the model and confidence scores; 2, constructing a related triad, an adjacent triad and entity long text description based on the query to form a context prompt, inputting the context prompt and the query into a large language model, and performing semantic reordering and scoring by the large language model; and 3, constructing a fine tuning data set, and performing fine tuning on the large language model to obtain the KGC task optimization-oriented large language model. And based on the KGC score, the LLM score and the dynamic weight, outputting a complementation result through joint prediction of a fusion result. According to the method, the structured reasoning ability of the small model and the deep semantic understanding of the large language are effectively combined, the prediction accuracy, robustness and specialty are remarkably improved, and the defects that a single model is weak in generalization ability and insufficient in semantic utilization are overcome.

The invention relates to the technical field of knowledge maps, and discloses a knowledge map completion method based on semantic-structure multi-level fusion. According to the technical scheme, for a training set positive example triple, a structured rationality score and a semantic correlation score are fused to screen high-quality training samples; in order to complement a query search relation path, screening a guiding path through a uniqueness index; integrating the query text, the entity description, the neighbor facts and the guide path to construct an enhanced input prompt; dynamic structure embedding is generated by using a relational graph convolutional network, the dynamic structure is mapped by an adapter module and then injected into a large language model, and completion prediction is completed by combining a fusion input fine tuning model. According to the method, through multi-level fusion of semantics and structural information, intelligent sample screening, prompt enhancement construction and structural information dynamic injection are achieved, the accuracy, reasoning ability and training efficiency of a large language model in a knowledge graph completion task are effectively improved, and the method is especially excellent in performance in complex relation reasoning and inductive completion scenes.

The invention discloses a document auditing method and device based on cooperation of a large model and a rule engine. The method comprises the following steps: analyzing an original document to form an intermediate representation file; extracting entities from the intermediate representation file to form an entity candidate set; inputting the thinking chain cue word into the large model to obtain a first triple, a first confidence coefficient and a first risk level, and calculating a first traceable score of the large model according to the reasoning path node; performing deterministic judgment by using the rule engine to obtain a second triple, a second confidence coefficient, a risk level and a second traceable score; calculating two weighted voting values and obtaining a conflict difference value; taking a conclusion corresponding to the high-weighted voting value as a final conclusion when the conflict difference value is smaller than a preset judgment threshold value; otherwise, starting an artificial rechecking process; and generating audit reports in various formats. According to the method, the advantages of relatively high language understanding ability of a large model and relatively high certainty of a rule engine are exerted, and the problem of missed checking or excessive marking is avoided.

The invention relates to the technical field of term libraries in the field of constructional engineering, and discloses a construction application method and system for a term library in the field of constructional engineering, and the method comprises the steps: preprocessing unstructured data to obtain structured constructional engineering text data, and extracting the structured constructional engineering text data based on a joint extraction architecture to obtain an extraction result; the method comprises the steps of obtaining structured data of terms and building engineering triple information, establishing a standard library and a term library according to the structured data, constructing a term-standard double-layer knowledge graph based on internal association of the standard library and the term library and the building engineering triple information, and responding to query of a user. And calling the standard library, the term library and the term-standard double-layer knowledge graph by a mixed decision-making mechanism based on rule reasoning and graph neural network learning to obtain a query result. According to the invention, the full-life-cycle intelligent closed loop of construction engineering terms from extraction, management to application is realized, and the problems of difficult acquisition of industry term knowledge, complex management and low application efficiency are solved.

The invention discloses a knowledge graph construction method and device based on a large language model and a judgment document, and the method comprises the steps: segmenting an original text of the judgment document, and obtaining a text fragment set; inputting the text fragment set into a pre-training language model and a large language model to obtain an entity set and a property clue set; performing cross-model consistency verification on the entity set and the property clue set to generate an entity-clue mapping table; performing disambiguation processing and standardization processing on entities in the entity set based on the entity-clue mapping table to obtain a standardized entity list; constructing an entity pair list based on the standardized entity list and extracting a relation triple to obtain a candidate triple set; and performing reverse verification and rule filtering on the candidate triple set, and constructing a target knowledge graph based on the high-confidence triple set. According to the method, the quality of the slave knowledge graph can be improved, and then the efficiency and accuracy of automatically discovering and associating debtor property clues from legal documents are improved.

The invention belongs to the technical field of knowledge graph construction, and particularly relates to a multi-source heterogeneous data fusion and knowledge graph automatic construction system, which comprises a semantic extraction module for extracting minimum semantic fragments from multi-source data and performing cross-source alignment to generate a symbolized semantic framework; the data purification module converts multi-source data into symbolic predicates according to the framework, filters low-evidence data and outputs a purification set; the ternary generation module inputs the initial triad candidate set into a minimum rule grammar, and extracts entities, relationships and attributes to generate an initial triad candidate set; the disambiguation calibration module generates entity two-hop topological fingerprints based on the candidate set, completes disambiguation alignment and corrects conflicts, and outputs unambiguous structured knowledge; the mapping fusion module fuses the mapping information with the minimum connected ontology, and constructs and verifies an initial mapping knowledge domain; and the incremental updating module processes newly added data, performs incremental merging and maintains consistency, and forms a complete knowledge graph. According to the method, through full-link automatic construction, efficient fusion and high-quality atlas are realized.

The invention relates to the field of data retrieval, and particularly provides a dynamic retrieval enhancement generation method and system based on a knowledge graph, a terminal and a medium, and the method comprises the following steps: extracting a structured triple from multi-source heterogeneous data through a large language model, and constructing a global knowledge graph by means of an entity linking technology; integrating a real-time data stream interface, and dynamically updating graph nodes and attributes based on an event-driven mechanism; adopting a RotatE model to respectively encode the entity and the relationship to a complex number space, and fusing to generate a mixed vector to construct an efficient index; after user query is received, topic nodes are positioned through semantic analysis, related entities are retrieved through mixed indexes, and multi-hop reasoning is executed along a relation path to generate reasoning sub-graphs and extended contexts; and finally, generating structured text answers by using a large language model, and adaptively outputting multi-modal results such as texts, charts and the like according to user requirements. According to the method, the knowledge updating timeliness, the complex query reasoning capability and the retrieval precision are effectively improved.

A system and computer-implemented method facilitate expansion of knowledge. The system allows for characterization of natural language documents and of search queries to locate those documents. A natural language processing (NLP) analyzer finds subject-verb-object (SVO) triplets in received text and assigns initial hierarchical classifications to word components of the SVO triplets. An SVO analyzer generates variation hierarchical classifications by varying the initial hierarchical classifications assigned, selects at least one hierarchical classification from the initial hierarchical classifications and variation hierarchical classifications, and produces a token stream of tokens. The tokens represent respective hierarchical classifications of the at least one hierarchical classification selected. The token stream produced may represent a natural language (NL) document to be stored to facilitate matching the NL document to a subsequently independently specified query. Alternatively, the token stream produced may represent a query and the token stream is used for generating a response to the query.

The invention discloses a complex system comprehensive multi-view consistency detection method based on a large model, and relates to the technical field of information system architecture design. The method comprises the steps that N view models to be detected are determined, and a rule base is constructed and formed; forming a triple set; performing rule retrieval on the triple set to obtain a corresponding rule subset; and constructing a complete prompt statement according to the prompt template and the rule subset, and reasoning the prompt statement by the large language model to obtain a consistency detection result. Based on the MBSE method, the ability of understanding, reasoning and applying knowledge in the field of system architecture and system engineering is remarkably improved, the automation degree, accuracy and robustness of multi-view consistency detection are improved, and the correctness, completeness and consistency of complex information system architecture design are guaranteed.

The invention relates to the technical field of knowledge maps, in particular to a few-sample knowledge map completion method and system fusing relation perception information bottleneck. The method comprises the following steps: S1, preprocessing an input triple; s2, building a global aggregation module, and updating entity embedding; s3, establishing a relationship aggregation module, and updating relationship embedding; s4, establishing a relationship-based information bottleneck module, filtering noise irrelevant to tasks, and meanwhile, retaining relationship-specific information; s5, building an EM attention pooling module, adaptively aggregating multi-path semantic representation, and highlighting the correlation between the entity and the relationship; and S6, establishing a score calculation module, calculating a triple score and outputting the triple score. The invention provides a few-sample knowledge graph completion method and a few-sample knowledge graph completion system fusing relation perception information bottleneck, which are used for solving the problems of insufficient relation and entity representation coupling, high redundant information interference and difficulty in modeling due to high-order relation dependence in a knowledge graph completion task, and realizing efficient inference of potential relation facts.

The invention provides a heterogeneous database structure dynamic synchronization and fault-tolerant migration method, system and device, and belongs to the technical field of database migration. The method comprises the following steps: detecting a structural difference through combination of active detection and passive triggering; realizing DDL statement conversion of different databases by utilizing AST analysis through a cross-platform DDL converter; defining a migration process by adopting a migration state machine model so as to ensure orderly migration; and a breakpoint resume and consistency guarantee mechanism is utilized to ensure that migration fault tolerance is consistent with data by means of a vernier triple, an intelligent recovery strategy and a bidirectional check protocol. According to the method, the problems of structure synchronization and migration fault tolerance of the heterogeneous database in an environment without native change log support are effectively solved, the automation, high efficiency, stability and reliability of database migration are improved, and the method has wide application value.

The invention provides a retrieval enhancement generation method and system in the dual-carbon field, and relates to the field of data processing. According to the method, multi-source unstructured data in the dual-carbon field is collected, after data preprocessing is carried out, a multi-granularity query problem set is formed, a dual-carbon field knowledge base is obtained, and a dual-carbon field-oriented embedding model CEMBING and a reordering model CReranker are constructed. The CEMBING model adopts a semantic partitioning method and a joint training strategy, so that semantic information of a dual-carbon field text can be effectively captured; the CReranker model adopts a negative example mining strategy and a triple loss function, candidate documents can be accurately sorted, the problems of knowledge limitation and insufficient timeliness of LLMs in the application of the dialogue system in the dual-carbon field are effectively solved, and the accuracy and efficiency of retrieval enhancement generation of the dialogue system in the dual-carbon field are improved.

The invention discloses a knowledge graph construction method, device and equipment and a readable storage medium, and is applied to the technical field of natural language processing and knowledge engineering.The method comprises the steps that document content is divided to obtain initial document fragments, and all the initial document fragments are merged and divided based on semantic similarity to obtain target division blocks; based on the target division block, subject-predicate-object formatting processing is carried out to obtain a subject-predicate-object formatting result; entity and relation extraction is carried out according to the subject-predicate-object formatting result to obtain a display triple, implicit relation reasoning is carried out to obtain an implicit relation triple, entity and relation type normalization is carried out to obtain a normalized triple, and the knowledge graph is constructed based on the normalized triple. The block segmentation driven by semantic similarity is adopted to avoid sentence breakage, cascade errors are reduced based on subject-object formatting processing, and the problems of fragmentation, link missing and the like are solved based on implicit relations, so that the integrity and accuracy of knowledge graph construction are improved.

The invention relates to the technical field of knowledge maps, and discloses a multi-modal knowledge map completion method based on a generative adversarial network, which is used for solving the problems in the prior art that modal information is unbalanced, a modal fusion strategy is rough, the inference capability is reduced under modal deficiency and the like. Particularly, when only a tail entity prediction task is carried out, good generalization ability and robustness are achieved, a pre-training comparison learning model is used for coding image and text information of an entity, and multi-modal semantic features are extracted and mapped to a unified embedding space to achieve modal alignment; secondly, through a fine-grained modal attention fusion module, dynamically adjusting weights according to contributions of different modals in a tail entity prediction task, and realizing effective integration of information; further introducing a generative adversarial network based on a convolutional structure, and enhancing the robustness of the model to modal deficiency by constructing pseudo-modal features; and then a RotatE scoring function is adopted to model triple semantic rationality, and training optimization is carried out in combination with a self-adversarial negative sampling strategy.

An alignment score is generated for a test language model from input data including a number of triplet data structures. The method also includes identifying, responsive to the alignment score failing to satisfy a score threshold, a fail triplet data structure in the number of triplet data structures for which the evaluation score includes the indication of fail. A judge language model is executed on the fail triplet data structure to output a type of misalignment that the test language model produced when the test language model executed on the fail prompt. The judge language model is re-executed on a combination of the fail triplet data structure and the type of misalignment to output a cause of the fail response. An enhanced prompt is generated accordingly and then returned.

The invention relates to the technical field of entity relationships, and discloses a multi-feature fused entity relationship extraction optimization model construction method and system, and the method comprises the steps: obtaining an input text sequence and a relationship label set, and carrying out the coding processing, and obtaining a text embedding vector and a relationship embedding vector; calculating the similarity between the text embedding vector and the relation embedding vector, screening in combination with a similarity threshold to obtain a candidate relation subset, and fusing the relation embedding vector and the text embedding vector in the candidate relation subset to generate a relation enhancement vector; processing the relation enhancement vector through a mixed attention mechanism to obtain a semantic enhancement vector; constructing a three-dimensional marking matrix based on the semantic enhancement vector, and marking the three-dimensional marking matrix by adopting a diagonal marking strategy; and carrying out decoding processing on the three-dimensional mark matrix, and generating a structured entity relationship triple by adopting a constraint decoding strategy. The interactive fusion of the text and the relation label can be accurately processed in a unified coding space.

The invention belongs to the field of computer vision and artificial intelligence, and provides a scene graph generation method and system based on dual-dependence joint learning, and the method achieves the feature alignment through the cross attention operation of the flattening features of an input image and the query vectors of a subject and an object. Then, subject and object features are analyzed, high-confidence pairs are screened out, and the high-confidence pairs and predicate query vectors are processed in a decoder to output triple semantic features. And constructing a global association graph based on the features, updating node features by using an attention graph convolutional network, and finally predicting the category and bounding box of each triple by using a multi-layer perceptron to complete scene graph generation. According to the method, the generation accuracy and the relation context consistency are improved, the method is suitable for the fields of image understanding, intelligent monitoring and the like, and a complete system architecture solution is provided.

The invention provides a bridge operation and maintenance knowledge management and question answering method based on a knowledge graph and a large language model. The method comprises the following steps: performing semantic dicing and dynamic cue word generation on texts such as a design report and an overhaul report through a large language model, automatically extracting an entity-relationship-attribute triple, completing entity co-reference resolution and relationship combination, and constructing an initial knowledge graph; further utilizing a curiosity algorithm to continuously merge low-similarity new knowledge in a subsequent detection report to realize iterative growth of the atlas along with an operation and maintenance life cycle; forming a multi-level community by adopting Leiden clustering, generating an abstract pyramid, and enhancing macroscopic semantic expression; and finally, through a semantic embedding model and a GraphRAG technology, precise matching and traceable answering from natural language questions to graph knowledge points are realized. According to the method, the problems of knowledge fragmentation, update lag and illusion in the bridge operation and maintenance field are solved, and the accuracy and robustness of operation and maintenance knowledge questions and answers are improved.

The invention provides a ticket business full-life-cycle anti-counterfeiting and verification system based on block chain trusted triple mapping, and relates to the field of electric digital data processing. Comprising a trusted triple construction and cryptography binding module, a block chain hash anchoring and consensus evidence storage module, a trusted triple dynamic evolution and state transition module and a multi-level anti-counterfeiting verification and anomaly detection module, the block chain Hash anchoring and consensus evidence storage module is used for storing a trusted triple in a block chain in a tamper-proof manner, and the trusted triple dynamic evolution and state transition module is used for realizing state tracking and evolution management of ticket business in a full life cycle. The multi-level anti-counterfeiting verification and anomaly detection module is used for providing static identity verification, dynamic anti-counterfeiting vouchers and abnormal behavior detection capability; according to the system, the security problem in a ticketing system is solved by enhancing the security of the entity identifier and protecting state transition privacy.

The invention provides a question and answer method, system and device based on a water conservancy project knowledge graph and a medium, and belongs to the technical field of artificial intelligence. The method comprises the following steps: converting water conservancy project unstructured data into structured data through an algorithm to construct an atlas triple, and synchronously storing the atlas triple to ES and Neo4j; establishing an MCP service to access a custom atlas service, and configuring a plug-in to connect Neo4j and a retrieval service; matching fuzzy knowledge of the question in the ES is obtained, a Neo4j mode is retrieved to filter irrelevant nodes, and query nodes and relations are judged; a Cypher statement is generated and verified; iteratively calling a service to obtain node relation data, identifying a relation type, judging a dynamic relation according to a calling rule, and filtering to generate the node relation data; and finally listing entities, relationships and attributes, and feeding back a natural language result of the user. According to the method, the utilization rate, query accuracy and reuse efficiency of water conservancy knowledge are effectively improved, and efficient support is provided for water conservancy decision making.

The invention discloses a high-efficiency lightweight federal recommendation method, which comprises the following steps that: a server initializes an article continuous value embedding matrix and an article offset vector, obtains a binary article matrix through binarization processing and distributes the binary article matrix to a client; the client constructs a training triple through difficult negative sample mining based on local data, carries out local training by adopting a mixed loss function combining sorting loss, reconstruction loss and regularization loss, updates user parameters and calculates an article gradient; and the server aggregates the gradient update global parameters uploaded by the client. According to the method, the model expression ability is enhanced by introducing bias terms, the sorting performance is optimized by adopting a mixed loss function, the training efficiency is improved by combining difficult negative sample mining, the recommendation precision is remarkably improved on the premise that the federated learning privacy protection advantage is completely reserved, meanwhile, the client storage, communication and calculation expenses are greatly reduced, and the user experience is improved. The method is especially suitable for mobile terminal deployment with limited resources.