85 results about "Graph embedding" patented technology

This invention discloses a system for identifying and simulating key nodes in traffic networks based on deep reinforcement learning. Addressing the high computational complexity of key node search and the lack of dynamic verification in traffic networks, this invention employs the following system: a road network topology mapping module parses GIS map data and removes redundant information to generate a weighted topology map; a deep reinforcement learning identification module, based on a deep Q-network, maps node features to graph embedding information through graph representation learning, with the optimization objective of minimizing cumulative normalized connectivity, outputting the optimal key node sequence; and a traffic effect simulation verification module uses a traffic simulation platform to establish a realistic road network model, quantitatively analyzing the impact of key node failures on average travel time and waiting time. This invention features low time complexity, the ability to generalize from small synthetic networks to extremely large-scale real networks, and simulations have verified its high application value in actual traffic management and disaster prevention.

The application relates to a merging vector optimization method in a mobile cell-free scenario, wherein the method comprises the following steps: generating and dividing a data sample set based on a pre-established full connection CF mMIMO network architecture; performing graph construction and initial graph embedding on random channel state information in a training set to obtain an initial graph embedding vector; performing forward propagation calculation on the initial graph embedding vector in combination with a pre-established graph neural network to obtain a current round merging vector; determining an initial graph neural network model when a training bit error rate meets a first bit error rate requirement; testing a test set to obtain a test bit error rate; determining a final graph neural network model when the test bit error rate meets a second bit error rate requirement; and deploying a final full connection CF mMIMO network architecture and performing network scheduling. Thus, the problems of high calculation complexity, low real-time performance and low reliability in the related art are solved, the calculation complexity is reduced, and the robustness, scalability and system fairness are improved.

The application provides a kind of method and device for fast matching of similar house type based on multi-level score, obtains the house type drawing of the house type to be matched, obtains the minimum circumscribed rectangle of the house type drawing according to the house type drawing to be matched, the length-width ratio value of the minimum circumscribed rectangle of the house type drawing to be matched and the minimum circumscribed rectangle of the target house type is sorted and screened, to obtain the first matching result, based on the space type distinction space area of the house type drawing to be matched, generate space connected topology graph, calculate the house type dynamic line, calculate the house type dynamic line similarity based on the house type dynamic line, obtain the first weight result based on the preset weight and the dynamic line similarity, obtain the room topology graph based on the room topology relationship graph embedding vector, assign different attention weights to the neighbor nodes of each node of the room topology graph, and calculate the similarity between node representations through the attention weight, calculate the second weight result based on the preset weight, and recommend each decoration scheme according to the sorting result.

This invention provides a vehicle simulation speed correction method and system, relating to the field of vehicle intelligent dynamics modeling technology. The correction method includes the following steps: S1: Input and process simulation data output from the vehicle dynamics simulation model and corresponding real vehicle test data; S2: Construct a dynamic graph structure representing the interaction between state variables based on a preset physical coupling relationship of the vehicle powertrain; S3: Input the simulation data into a GCN and perform graph convolution operations under the constraints of the dynamic graph structure to extract graph embedding feature sequences representing the spatial dependencies between state variables; S4: Input the graph embedding feature sequences into a TCN and perform temporal convolution operations to learn the dynamic evolution law of state variables in the time dimension and output the correction amount of the vehicle simulation speed; S5: Output the result. Based on this, this invention solves the problem that existing correction methods have various limitations in practical applications.

This invention relates to the field of rural economic data analysis technology, specifically an artificial intelligence-based method for rural economic data analysis. The method includes: delineating an economic data perception domain in a geographic information system based on the administrative division vector data of the target administrative village; collecting multi-source heterogeneous economic data through a protocol parser of an agricultural IoT terminal and converting it into standardized economic behavior data; extracting entities using natural language processing technology to generate a structured rural economic behavior fact table; inputting the fact table into a pre-trained economic activity association model and constructing a rural industry relationship graph using a graph embedding algorithm; and calculating the centrality index of economic entities based on the graph topology to identify core operating entities. This method enables accurate processing and association analysis of rural economic data, clearly presenting industry relationships and providing effective technical support for rural economic analysis.

This application relates to the field of integrated circuit design automation technology, and provides a method and system for verifying analog circuit structures based on graph neural networks. The method includes: converting a SPICE netlist into a graph structure representation, where nodes correspond to devices and edges correspond to electrical connections, and extracting feature vectors containing device type, parameters, functional roles, connection types, and signal directions; encoding the circuit diagram using a graph attention network to obtain node embeddings and graph embeddings; performing multi-task prediction based on the embedding results, including node classification, graph classification, structure scoring, and error localization; and finally generating a verification report containing an error list, location information, and repair suggestions. This application achieves automatic detection and precise localization of errors in analog circuit structures, improving detection accuracy by a significant margin and possessing strong generalization capabilities.

The application provides a few-shot graph-level anomaly detection method based on structure perception prompt, comprising: pre-training a detection model to obtain a pre-trained detection model; fine-tuning the detection model by using labeled graphs to obtain a fine-tuned detection model, wherein the labeled graphs comprise labeled normal graphs and labeled abnormal graphs. The model pre-training comprises: constructing an enhanced graph set and a sampled graph set for an original graph set; constructing a graph-graph structure for the original graph set, the enhanced graph set and the sampled graph set; performing message propagation on the graph-level representation on the graph-graph structure to generate a graph embedding matrix; and performing joint optimization of node-graph contrast learning and graph-graph contrast learning on the detection model by using the graph-level representation and the graph embedding matrix to obtain the pre-trained detection model. The application effectively improves the detection accuracy and generalization ability of the model.

This invention discloses a target tracking method and system based on UAV ID tracing, relating to the field of low-altitude security technology. The method includes: performing target detection on the current frame image to obtain detection boxes and detection confidence levels; simultaneously receiving UAV Remote ID information and extracting radio frequency fingerprint features to construct a multi-factor identity identifier; constructing a local geometric neighborhood for the detected target, calculating relative geometric relationships, and encoding them as geometric feature vectors; expanding the spatial graph of consecutive multiple frames into a spatiotemporal heterogeneous graph, embedding the multi-factor identity identifier, and outputting an association confidence matrix through a spatiotemporal graph neural network; implementing a three-level cascaded association based on dynamic thresholds, including high-confidence deterministic matching, medium-confidence fine association, and re-identification comparison of unmatched targets; updating the trajectory state and outputting a trajectory sequence with identity identifiers for flight tracing. This invention solves the problems of UAV identity hopping, occlusion loss, and ID spoofing in complex environments.

This disclosure relates to the field of computer technology and discloses a method, apparatus, computer device, and storage medium for identifying associated communities. The method includes: acquiring each node in a social network and the atomic relationships between them; for any target node among the nodes, determining the graph embedding information corresponding to the target node based on the atomic relationships; constructing a social graph between the target node and all other nodes according to the graph embedding information; and identifying the target associated community from the social graph based on the node relationships in the social graph. By implementing the technical solution of this disclosure, multi-path information between nodes can be considered, facilitating better capture of complex relationships and features between nodes, and fully utilizing the graph data structure information of the social graph to achieve the identification of associated communities.

The application belongs to the technical field of artificial intelligence and intelligent education, and discloses a knowledge graph joint decoding method and system based on coverage planning and backtracking navigation for education content generation, which is used for solving the problems of incomplete knowledge coverage and disorder of teaching preposition dependency in automatic generation of education content. The method contains three cooperative mechanisms: pre-decoding coverage planning, learning order constraint topological sorting based on an education knowledge graph to generate a teaching planning sequence and allocate a length budget; double-head joint decoding, maintaining a graph cursor, and outputting teaching text and navigation actions through the cooperation of a word prediction head and a navigation prediction head, and fusing graph embedding through a gating mechanism; preposition dependency backtracking, automatically inserting teaching review text and then restoring when a preposition knowledge reference intention is not covered, and simulating instant knowledge connection in teacher teaching. The knowledge coverage rate of the method is 98.3%, and the preposition dependency violation rate is less than 0.5%.

This invention relates to the field of secure communication application technology, specifically to a host asset identification and profiling method based on multimodal network feature fusion enhancement. First, a data collection agent is deployed at the core node of the target network to collect multimodal data through traffic mirroring and log crawling, and a cross-modal aligned multi-source network data representation is constructed. Next, a TCN-BiLSTM model combined with an attention mechanism is used to extract deep periodic features, and a GraphSAGE model is used to generate graph embedding vectors bound to topological relationships. A pre-trained BERT model is used to obtain global protocol semantic feature vectors and communication intent distribution. Subsequently, a modality-specific mapping is used to unify dimensions, and a two-layer mechanism of intra-modal self-attention and inter-modal mutual attention is constructed to dynamically allocate weights, fusing multimodal features to extract core features of host assets. Finally, the fused features are mapped to structured text, and a CLM strategy combined with LoRA technology is used to fine-tune the LLM, generating a natural language asset profile containing role positioning, behavioral patterns, and topological relationship logic.

The application discloses a contract attack detection method and device based on symbolic execution and a graph neural network, comprising: obtaining deployment bytecode and runtime bytecode of a smart contract; performing symbolic execution on the runtime bytecode, extracting external calls and event information through depth-first search and taint tracking; constructing a contract call information graph and iteratively simplifying it; performing multi-dimensional feature coding on each node to generate a node feature vector; encoding on-chain metadata of a deployer into a deployer feature vector; constructing and training a graph embedding network and a classification network, the input of the graph embedding network being the simplified contract call information graph, the node feature vector and edge type information, and the output being a graph-level feature vector; the input of the classification network being a joint feature vector obtained by splicing the graph-level feature vector and the deployer feature vector, and the output being a contract classification result. The application can cover multiple attack types and realize attack contract identification with low false alarm rate by extracting contract execution logic structures for detection.

Embodiments of the present specification provide a method and device for training a graph neural network. In the method, a user relationship network graph is obtained. The user relationship network graph is divided into multiple subgraphs. Each subgraph is processed by a graph embedding using a graph neural network to obtain a user representation of each user node. A prediction network is used to obtain a prediction result for a target service based on the user representation. The first parameter of the graph neural network and the second parameter of the prediction network are adjusted to minimize the objective value under a target constraint condition. The target constraint condition is that, for each subgraph, the subgraph loss obtained using the adjusted second parameter is minimized compared to other second parameters. The subgraph loss is the sum of user losses calculated using a preset loss function on the prediction result and the business label of each user node in the subgraph. The objective value is the sum of the subgraph losses of each subgraph obtained using the adjusted second parameter.

This invention discloses a method for detecting quality faults in a flotation process based on distributed dynamic graph embedding. Step 1: Collect process variables such as concentration, pH value, ore fineness, froth layer thickness, and maximum allowable density during the flotation production process as input variables, and the concentrate grade of the flotation process as the output quality indicator. Step 2: Based on the mutual information method, establish a cross-correlation matrix MI between process variables and quality variables, calculate the average threshold, and select key variables. Step 3: Based on the production process and field experience of the flotation process, divide the key variables into Q quality-related sub-blocks and B quality-unrelated sub-blocks. Step 4: Use the key variable selection and sub-block decomposition results to establish a distributed dynamic graph model to achieve quality-related fault detection. Step 5: Use a Bayesian fusion network to fuse the monitoring results for decision-making. Step 6: Determine whether the fault is quality-related based on the monitoring results.

The application discloses a kind of vehicle networking information age optimization method and system based on graph reinforcement learning, first constructs the batch modeling of vehicle state and limited buffer queue structure, and the perception data is modeled as multiple data packet batches and is managed in queue;Graph neural network is used to model V2V link topology, and large-scale channel embedding representation reflecting topological structure is extracted;A multi-agent reinforcement learning system based on centralized training and distributed execution framework is constructed, each agent makes decisions according to the local state containing graph embedding features, outputs a hybrid action space, and receives the inverse number of end AoI as a reward;Introduce the graph embedding supervision mechanism based on advantage function, align the topological features with the long-term optimization goal;Through multiple rounds of training to update network parameters, finally realize the autonomous optimization control of each agent to packet loss and power.The application can effectively coordinate packet queue management and wireless resource allocation, and realize efficient and low-overhead AoI minimization in complex dynamic topology environment.

This invention discloses a supplier cheating detection method based on encoding / decoding algorithms, comprising: collecting node data information; calculating correlation using mutual information, and selecting combinations based on correlation thresholds to obtain combined data features; constructing a graph structure, training a bandwidth prediction model using GCN on the graph structure, optimizing the model using the mean square error between the model output and the actual bandwidth traffic as a first loss function, and obtaining a graph embedding representation based on the optimized bandwidth prediction model; concatenating the graph embedding representation, data information, and combined data features to obtain abnormal prediction demand data; dividing the dataset based on business and corresponding loss rate thresholds, updating the overall loss function of the abnormal detection algorithm using the first loss function to obtain an abnormal detection model; using the abnormal detection model for detection, manually confirming the detection results for false positives, and processing suppliers or optimizing the abnormal detection model based on the false positive results. This invention can cope with diverse cheating behaviors and achieve efficient cheating detection.

This invention discloses a human resource intelligent analysis method based on machine learning using multi-source data, relating to the fields of human resource management and artificial intelligence. The method includes: acquiring employee static profiles and individual historical data; generating individual basic potential vectors through machine learning model encoding; constructing a multimodal organizational network graph integrating organizational structure and collaborative relationships; applying a graph embedding model to learn and generate network fusion latent vectors, and grouping employees accordingly; collecting temporal interaction metadata for specific groups, inputting it into a temporal analysis model to construct a dynamic knowledge influence profile; and collecting skill acquisition data and organizational needs, generating personalized skill growth prediction maps through a temporal prediction model. This invention, by integrating multi-source data and revealing implicit network structures, achieves dynamic, multi-dimensional, and forward-looking assessment of talent, improving the scientific rigor and accuracy of human resource decision-making.

The application relates to the technical field of intelligent networked vehicles, and proposes an attack tracing method for vehicle-mounted intrusion detection, an electronic device and a computer program product. The method comprises the following steps: when it is determined that a vehicle is subjected to an attack behavior according to current communication data of the vehicle, a knowledge graph constructed based on the current communication data is acquired, the knowledge graph has obtained a node embedding vector of each node and an edge embedding vector of each edge through graph embedding processing; according to the node embedding vector and the edge embedding vector, an abnormal score and a centrality score of each node in the knowledge graph are respectively determined, and an abnormal score of each edge in the knowledge graph is respectively determined; and according to the abnormal score and the centrality score of each node in the knowledge graph and the abnormal score of each edge in the knowledge graph, an attack path of the attack behavior is determined. By using the method, the specific attack path can be traced back when it is determined that the vehicle is subjected to the attack behavior, so that the interpretability of the intrusion detection result is improved.

The application discloses a patent matching method based on large model and knowledge graph fusion, comprising the following steps: obtaining a semantic vector by a pre-trained large model according to collected patent text data, obtaining a graph embedding vector by a pre-trained knowledge graph model, and fusing the semantic vector and the graph embedding vector into a joint representation vector; further comprising: obtaining a semantic vector by a pre-trained large model according to collected user demand data, obtaining a graph embedding vector by a pre-trained knowledge graph model, and obtaining a joint representation vector; performing patent preliminary screening according to the vector similarity between the joint representation vector of the user demand data and the patent text data, obtaining a correlation score by comparing the preliminary screened patent text data and the user demand data, fusing the vector similarity through the correlation score, and obtaining a patent recommendation sequence. The core problem of poor matching precision caused by shallow semantic understanding and insufficient knowledge utilization is solved.

The application discloses a CircRNA-miRNA correlation prediction method and device and a medium. The method comprises the following steps: constructing a data set, wherein the data set comprises a plurality of CircRNAs and a plurality of miRNAs; based on the data set, biological attribute features are extracted from word embeddings of CircRNA sequences and miRNA sequences; behavior features are extracted based on isomorphic graph embedding; training features of a fusion carrier are obtained based on the biological attribute features and the behavior features, so that the interaction scores of circRNAs and miRNAs are learned, and potential CMAs are inferred. The application successfully predicts the complex relationship between circRNAs and miRNAs, the accuracy is 82.90%, and the AUC is 0.9075.

This application discloses a real-time decision-making method, system, and medium for intelligent agents based on Riemannian manifolds, relating to the field of artificial intelligence technology. The method, executed by an intelligent decision-making system, includes: constructing a behavior transition graph based on a historical trajectory dataset; mapping the attention matrix generated by a large language model in the decision-making task to a symmetric positive definite matrix manifold for representation; embedding the behavior transition graph into a hyperbolic Riemannian space for representation; learning the geometric alignment mapping from the symmetric positive definite matrix manifold and hyperbolic Riemannian space to a common metric space by optimizing the objective loss function; in the inference phase, acquiring current environmental observation data and inputting it into the large language model to obtain the corresponding attention representation, which is then mapped to the common metric space to obtain a query vector; and using actions corresponding to action nodes with a geometric distance below a preset threshold as control commands. This significantly reduces the computational overhead of decision-making and improves cross-scenario generalization ability and decision interpretability.

The application relates to a vehicle path optimization method and system based on deep reinforcement learning, which comprises the following steps: S1, acquiring node feature embedding based on a problem instance; S2, acquiring position feature embedding based on the problem instance; S3, using a multilayer perception machine to fuse the node and position feature embedding information and completing coding; S4, using a removal decoder to realize removal operation of a PDP node pair; S5, using a repair decoder to realize reinsertion operation of the removed node pair; S6, performing corresponding actions according to the output of the removal and repair decoders to realize state transition and continuously iteratively improving a scheme; S7, using a dynamic mode to update a graph embedding in real time according to the output of the decoder; and S8, training a model by using a proximal policy optimization algorithm and finally outputting a path optimization sequence. The application combines deep reinforcement learning and a heuristic method, trains a neural network to guide a local search algorithm, learns an optimization strategy in a step-by-step manner and iteratively improves a solution.

The application discloses a kind of based on multi-layer graph embedding atomization component intelligent mapping and process reconstruction method, system, equipment and medium, belong to component reconstruction technical field, comprising: component feature matrix is constructed by multi-source feature extraction, and based on this, multi-layer dependency graph is established to learn component embedding representation;Similarity matching and mapping are carried out using embedding, and the mapping result is optimized in combination with dependency consistency constraint, and then the process dependency topology is constructed and optimized to eliminate redundancy and cycle;Adaptively update dependency graph by detecting historical changes, and introduce expert knowledge constraint to correct and jointly optimize the result, and feed back the revised knowledge to knowledge base.The application realizes unified modeling and deep fusion of component structure, semantics and execution layer dependency relationship by multi-layer graph embedding technology, improves the precision and panoramic understanding ability of component mapping;Ensure the timeliness and accuracy of mapping and reconstruction result.

The application discloses a wind power cluster power prediction method based on global information graph fusion continuous learning, and the method comprises the following steps: step one, calculating the space-time correlation coefficient; step two, generating a dynamic adjacency matrix; step three, graph embedding based on Laplace mapping; step four, extracting the features of the embedded vector through a clustering algorithm; step five, constructing a global information graph; step six, denoising the global information graph by using an adaptive time sequence label smoothing method; step seven, establishing a graph attention network model of the fusion continuous learning strategy; and step eight, model training and testing. The application can accurately predict the power generation of large wind power clusters, and helps to improve the economy and safety of the operation of the wind power field. Meanwhile, the method provides reliable data support for power grid dispatching and wind power consumption, guarantees the safe and stable operation of the power system, and promotes the high proportion of renewable energy access and the construction of a green and low-carbon energy system.

A pipeline has been created that leverages artificial intelligence and machine learning to efficiently extract information from CTI reports obtained from various sources and yielding information that assists security analysts / threat teams (e.g., security operations centers (SoCs)) and improving the quality of CTI. The “CTI analysis pipeline” employs generative artificial intelligence (“genAI”) to summarize a collection of CTI threat reports and extract threat-related information including TTPs from the CTI reports. Relationships among the threat reports are determined based on the extracted threat-related information and encoded in a graph structure. Graph embeddings based on the relationships encoded in the graph structure and semantic embeddings from the report summaries are combined and the combined embeddings are clustered. The resulting clusters and trained clustering model can be used in various ways to improve CTI, such as determining malicious campaigns, augmenting existing campaign information, and detecting new IOCs and TTPs for existing campaigns and new campaigns.

This invention discloses a method for topological hierarchical knowledge detection in graph model representation learning. The method includes: acquiring graph embedding models based on various types of graph neural networks for different datasets to extract graph structure information, including node information and graph embedding representations; pre-training the graph embedding models based on a set loss target for different downstream tasks, and during training, using different types of probes to detect the representational capabilities of the graph embedding models as detection results, including centrality probes and distance probes. The centrality probe is used to evaluate the embedding quality of the embedding vectors to centrality information, and the distance probe is used to evaluate the embedding amount of the embedding vectors to topological distance information; and selecting the type of graph embedding model based on the detection results for the target downstream task. This invention can reveal whether node influence is encoded in graph representation learning.

This invention discloses a factory production scheduling system and method based on real-time data and AI decision-making, relating to the fields of intelligent manufacturing and production control technology. It constructs a dynamic topology map of the workshop, including resources, tasks, and buffer nodes, by collecting multi-source heterogeneous industrial data. A multi-layer convolutional network is used to extract the deep features of the map, generating a high-dimensional graph embedding feature tensor containing global spatiotemporal dependencies. A binary action mask vector is generated by combining physical and process constraints. The feature tensor is input into a DQN neural network to generate an original value score, and the mask vector is used to correct and obtain the optimal legal scheduling action. This invention achieves global perception of workshop congestion and fault propagation through a multi-layer convolutional network, ensures the physical feasibility of scheduling instructions through action masks, and achieves coordinated optimization of production capacity, energy consumption, and equipment health through multi-objective reinforcement learning, significantly improving the factory's adaptive scheduling capability under dynamic disturbance environments.

The application provides a wind turbine operation and maintenance management method based on a knowledge graph, comprising the following steps: performing text content and feature analysis and preprocessing, dividing entity types according to the similarities and differences of entity properties, giving constraint definitions for the relationships of entity types at various levels, and extracting entity words for wind farm operation and maintenance by using a rule matching algorithm based on an improved TextRank; defining the relationships between various entities according to field texts, and realizing the relationship linkage between different entities by using an entity relationship extraction algorithm based on rule matching; constructing a safety management procedure graph and a device operation and maintenance requirement graph established based on a protégé wind power equipment ontology instance topological relationship library; realizing the visualization of the knowledge graph for wind turbine operation and maintenance management, and forming an application scheme by using a query language and a graph embedding technology. Compared with the prior art, the application has the advantages of high accuracy, high efficiency and strong field professional application.