2179 results about "Cluster result" patented technology

The invention discloses an APT attack path reconstruction method based on time sequence diagram comparison clustering and a medium. A security event standardized data set is obtained; mapping each security event into a multi-modal node through a heterogeneous time sequence diagram set construction method, and generating a directed edge to construct and complete a heterogeneous time sequence diagram set; a stage embedding time sequence diagram set is obtained through the joint attack stage set; outputting a time sequence diagram similarity matrix by adopting a multi-scale diagram similarity algorithm of time alignment perception; generating an event semantic sparse matrix based on the threat intelligence knowledge graph; obtaining an image clustering result set; uncertain samples in the graph clustering result set are obtained and processed, and an APT attack path reconstruction result is obtained. The problem that an APT attack path reconstruction method mainly depends on rule matching, single-dimensional feature comparison and manual analysis of security logs or alarm streams is solved, the interpretability of APT traceability is greatly enhanced, and subjective errors of manual research and judgment are reduced.

The invention is suitable for the technical field of agricultural intellectualization, and provides an intelligent blueberry disease detection method and system based on multi-modal unsupervised learning, and the method comprises the following steps: carrying out the feature extraction and clustering of preprocessed multi-modal data based on an adaptive contrast deep clustering framework, and obtaining a feature extraction result; obtaining a multi-modal preliminary feature and a preliminary clustering result; based on a multi-modal complementary feature fusion mechanism, according to the multi-modal preliminary features and the preliminary clustering result, carrying out adaptive weighted fusion on the multi-modal preliminary features to obtain multi-modal fusion features; performing unsupervised clustering optimization and disease type identification on the multi-modal fusion features to obtain an unsupervised learning model; and performing deployment and incremental learning on the unsupervised learning model, and detecting the blueberry diseases. According to the method, early-stage accurate detection of blueberry diseases is realized through an unsupervised learning algorithm, a new normal form is provided for intelligent accurate management of blueberries, and the disease prevention and control efficiency and industrial economic benefits are effectively improved.

The invention provides a low-orbit satellite resource allocation method and device based on beam and frequency domain collaborative optimization, and the method comprises the steps: dividing the overall bandwidth of a user region into a central sub-band and an edge sub-band through employing a dynamic soft frequency reuse scheme; all ground users are subjected to spatial clustering according to geographic positions and traffic demands, coverage cells are dynamically divided according to clustering results, and balanced service is achieved; on the basis of dynamic cell division, user clusters with high service volume are selected according to priorities, a beam activation set is generated, and a hopping beam matrix is updated; establishing a channel model between a user and a beam and a switching matrix between satellites; constructing a system utility target, and constructing a multi-target optimization problem model by taking the system utility target as a main part and integrating interference, switching overhead and delay penalty; and the multi-dimensional resource collaborative optimization framework based on deep reinforcement learning reconstructs a multi-objective optimization problem model, and outputs a solution of the multi-objective optimization problem model, including an optimal bandwidth and a power allocation result. According to the invention, resource allocation can be carried out on low-orbit satellites.

The invention relates to the technical field of fault monitoring, in particular to a power transmission line fault monitoring and positioning method, system and device and a medium. The method comprises the following steps: firstly, performing digital conversion on an original current signal to obtain time sequence data, and then identifying and classifying fault events through threshold detection and a feature extraction algorithm to obtain classification labels of fault types; then, based on the classification label, establishing a traveling wave propagation model and carrying out time difference calculation, carrying out clustering analysis on a plurality of wave head data in a preset time window, and judging whether the fault is a single-point fault or a multi-point fault through a clustering result; and finally, a positioning result is transmitted to the central server for comprehensive verification and alarm management, and a final fault response is formed. By introducing a time series data analysis and clustering judgment mechanism, the fault type can be accurately identified, single-point and multi-point fault conditions can be effectively distinguished, and meanwhile, the reliability of a positioning result is improved through comprehensive verification of the central server.

The invention relates to the technical field of data analysis, in particular to a diabetes clinical test data intelligent clustering analysis system and method based on federal learning. Comprising a data acquisition and preprocessing unit; a federal privacy protection unit; the dynamic clustering modeling unit is used for constructing a clustering model of self-adaptive diabetes data features, and realizing joint clustering analysis of multi-source heterogeneous data by adopting a hierarchical federal architecture and a dynamic parameter aggregation algorithm and combining a diabetes course time decay factor and a clinical feature weight adjustment strategy; a double-track verification optimization unit; and an intelligent decision support unit. According to the method, the incidence matrix of the diabetes disease course time decay factor and the clinical characteristics is introduced, the dynamic weight vector is constructed and applied to clustering distance calculation, so that the model can adapt to dynamic changes of the clinical characteristics in different disease course stages, the adaptability to multi-center heterogeneous data is improved, and the stability of a clustering result is enhanced.

The invention discloses a marketing channel clustering modeling and optimizing method driven by heterogeneous data. The method comprises the steps of obtaining a multi-dimensional data set of a marketing channel, and performing preprocessing; the data set comprises structured data and unstructured data; and extracting a text word segmentation result and an emotion feature from the unstructured data, and extracting an image feature to generate a structured feature set, an unstructured feature set and the like. According to the method, structured and unstructured data can be integrated, and potential information of marketing channels can be comprehensively mined. Through a deep learning model (an auto-encoder, a variational auto-encoder and a generative adversarial network), features are extracted and optimized layer by layer, and the quality of the features and the performance of the model are improved. And in combination with a clustering algorithm and an evaluation index, the accuracy and reliability of a clustering result are ensured, and powerful support is provided for optimization of a marketing channel.

The invention relates to the field of data processing, and discloses an intelligent data processing system based on an AI large model, which comprises the following steps: when detecting that a plurality of candidate entities exist in a text, judging whether the candidate entities need to be subjected to semantic disambiguation processing or not according to context semantic relevancy and entity historical reference frequency; constructing a cross-domain context representation vector, introducing a pre-training language model to encode the context, generating deep semantic representation of candidate entities, and judging whether a clustering result has ambiguity or not; whether the candidate entities have conflicts or not is judged by fusing language model embedding and structuring knowledge graph information; constructing an entity relationship graph based on a graph neural network, performing causality, temporal and attribute dependence reasoning on the relationship between the existing candidate entities, and judging whether the relationship between the candidate entities should be combined or split; and carrying out label replacement on the candidate entities in the target text in combination with the context and the reasoning result. The method has the advantage of improving the accuracy of data processing.

The invention provides a pseudo-tag-based intention recognition model training method, an intention recognition method and an intention recognition device. The method comprises the following steps: inputting a sample text into a language model to extract a feature vector; clustering the sample text based on the feature vector, taking a clustering result as a pseudo tag, and calculating normalized mutual information of the real tag and the pseudo tag of the labeled sample text; determining a confidence score corresponding to each sample; the confidence score is used for quantifying noise in the pseudo tag, screening a high-confidence sample and taking the corresponding pseudo tag as a self-supervision signal, and iteratively optimizing the language model until convergence; after iteration, clustering is initialized again, a clustering result is updated, and mutual information and the number of iterations are normalized; when the number of iterations reaches an upper limit or the normalized mutual information amplification is smaller than a threshold value, training is terminated, and the language model is determined as an intention recognition model; the problem that the new intention recognition capability of the model is reduced due to continuous propagation and accumulation of noise pseudo labels can be solved; and the new intention recognition capability of the model is improved.

A point cloud cylinder segmentation processing method and apparatus, and an electronic device. The present invention relates to the technical field of computers, and in particular, to the technical field of three-dimensional point cloud segmentation. The method comprises: acquiring a plurality of sample points comprised in a point cloud corresponding to a target object; determining neighbor points of the plurality of sample points within respective preset neighborhood ranges; on the basis of the neighbor points of the plurality of sample points within respective preset neighborhood ranges, obtaining normal vectors respectively corresponding to the plurality of sample points; performing clustering processing on the plurality of sample points on the basis of the normal vectors respectively corresponding to the plurality of sample points to obtain a clustering result; and performing cylinder segmentation on the target object on the basis of the clustering result to obtain a cylinder segmentation result of the target object.

The invention discloses a composite material wind power blade damage detection method, and aims to solve the problems of difficulty in quantitative analysis and low distinguishing precision of various damage types in composite material wind power blade damage detection in the prior art. Comprising the following steps that acoustic emission signals in the operation process of the wind power blade are obtained in real time and preprocessed, and the acoustic emission signals comprise parameter data and waveform data; performing unsupervised clustering on the preprocessed parameter data to obtain a preliminary clustering result; training the preprocessed waveform data by adopting an improved multi-branch convolutional neural network, and outputting an accurate classification result; the unsupervised clustering result is verified by using the supervised learning classification result, and the accuracy of the clustering result is evaluated; and fusing the quantitative index accumulated energy, the damage type and the frequency, and constructing a comprehensive damage evaluation model to determine the damage degree. According to the method, the unsupervised clustering method and the supervised learning method are combined, and accurate identification and quantification of the damage are realized.

The invention discloses an unknown radar radiation source sorting method based on time sequence feature clustering. The unknown radar radiation source sorting method comprises the following steps: acquiring radar signal pulse description words of pulse signals; according to the radar signal pulse description word, performing spatial clustering on the pulse signal to obtain a spatial clustering result; for each spatial clustering result, according to the radar signal pulse description word of the pulse signal in each spatial clustering result, performing time feature clustering on the pulse signal in each spatial clustering result to obtain a time sequence clustering result; and analyzing a time parallel relationship and a time continuous relationship of the time sequence clustering results, and performing pulse group sequence blending on the time sequence clustering results to obtain a sorting result. On the basis of the existing clustering algorithm, the information of the pulse signal in the time dimension is introduced, the time sequence feature clustering of the pulse signal is realized by using the ST-DBSCAN algorithm thought and introducing the TOA-PA constraint interval, and the method has strong robustness for the complex electromagnetic environment.

The invention discloses an industry load prediction method, system and device fusing standard mutual information and improved bidirectional LSTM, and relates to the technical field of power systems. At present, industry monthly load prediction is not accurate. The method comprises the following steps: carrying out clustering processing on information including a load sequence, a consumption level, an air temperature and vacation by adopting an improved fuzzy C-means clustering algorithm considering kernel density estimation; standard mutual information calculation is carried out based on a clustering result, and quantitative analysis is carried out on relevance between information factors and industry monthly loads; the method comprises the following steps of: selecting an industry load, distributing weights, constructing a bidirectional LSTM neural network to capture a time sequence change rule of the industry load, analyzing external feature influence by adopting a multi-head attention mechanism, and outputting an industry monthly load prediction result through convolutional neural network connection. According to the technical scheme, the influence of external influence factors on the industry monthly load can be effectively considered, and the accuracy of an industry monthly load prediction result is improved.

The invention discloses an online video content intelligent pushing method combined with a learning interest model. The method comprises the following steps: constructing a dynamic interest vector based on multi-source user behavior data, generating a user interest portrait vector set, and performing interest dimension clustering and weight distribution; generating a video content feature vector set according to a clustering result of the user interest portrait vector set; establishing a multi-dimensional association relationship between the user interest portrait vector set and the video content feature vector set, and outputting a user-video matching confidence matrix; converting the user-video matching confidence coefficient matrix into a push sequence based on a multi-objective optimization strategy and issuing the push sequence; and feedback behaviors of the user on the pushed video are collected in real time to realize closed-loop optimization. The method has the following advantages and effects: accurate perception and deep semantic matching of the dynamic learning interest of the user can be realized, and the accuracy, timeliness and user satisfaction of content distribution are remarkably improved, so that the learning efficiency and experience are optimized.

Methods and systems for interacting with users via a chatbot. A natural language query is received and processed by submitting a search query to a search engine. The search engine identifies relevant information including textual information and images for formulating a response. The identified information and query are submitted to a Large Language Model which generates a response displayed via the chatbot. The response may include textual information and relevant images. The system can extract text from images of documents and convert textual information into numerical vector representations for processing. Selectable options based on clustered relevant information can be provided to users for query refinement when appropriate. The chatbot interface enables natural language interactions while leveraging search capabilities and Artificial Intelligence to provide informative and helpful responses with both text and visual elements.

The invention discloses an event alarm method, device and equipment, a medium and a product. The event alarm method comprises the following steps: receiving an alarm information set sent by each alarm source; preprocessing the alarm information set to obtain a processed alarm data set; performing similarity analysis on the alarm data items in the alarm data set to obtain a similarity data set among the alarm data items; and according to an association clustering algorithm and the similarity data set, determining clustering results of different types of alarm events, and giving an alarm. According to the method, similarity analysis is carried out on a received original alarm information set, a correlation clustering algorithm is adopted to further reduce the calculation amount and more accurately identify the similarity between the alarm information, and the similar alarm information is gathered together to form a representative clustering result, so that the accuracy of alarm information clustering is improved. And operation and maintenance personnel can carry out troubleshooting and root cause analysis more easily.

The invention belongs to the field of pedestrian re-identification, and relates to a video-based unsupervised visible light infrared pedestrian re-identification method, which comprises the following steps: acquiring query data and a data set, inputting the query data and the data set into a trained re-identification model to obtain query features and a feature set, and matching the query features with the feature set to obtain an identification result; the training process of the re-identification model comprises the following steps: acquiring visible light data SV and infrared data ST; inputting the SV and the ST into a feature extraction module to obtain visible light and infrared features FV and FT; inputting the FV and the FT into a clustering module to obtain a clustering result; inputting the clustering result into a progressive false label correction module to obtain a corrected clustering result; inputting the SV and the ST into a feature extraction module to obtain visible light and infrared features qV and qT; updating model parameters according to the qV, the qT and the corrected clustering result until a trained re-identification model is obtained; according to the method, noise samples are recovered into effective labels through intra-modal correction and inter-modal correction, and robustness is enhanced.

The invention provides a large-scale encrypted traffic frame-by-frame clustering analysis method based on a big data architecture, and relates to the technical field of big data, and the method comprises the steps: carrying out the data partitioning and storage of target encrypted data obtained through the preprocessing of original encrypted traffic data; based on a clustering visualization result obtained by visualizing a target clustering result obtained by carrying out frame clustering analysis on the target encrypted data, judging whether a data traffic abnormal behavior exists or not, and when the data traffic abnormal behavior exists, generating an abnormal analysis report; and an abnormal analysis report is transmitted to safety management personnel so as to take corresponding measures in time for defense. The method comprises the following steps: visualizing a target clustering result obtained by carrying out frame clustering analysis after processing and partition storage on encrypted traffic data based on a big data architecture, identifying traffic data exception, generating an exception analysis report when the exception exists, and transmitting the exception analysis report to safety management personnel to take corresponding measures for defense. The network threat identification capability is effectively enhanced, and the overall protection level of network security is further improved.

The invention discloses a fault diagnosis method and a fault diagnosis system for an electric valve actuating mechanism. The method comprises the following steps: synchronously acquiring signals through an anti-EMI (Electro-Magnetic Interference) multi-source sensor; adopting complex Morlet wavelet packet decomposition to extract a 1.2-2.4 kHz energy entropy minimum frequency band, and calculating a kurtosis index; separating the third harmonic of the current through variational mode decomposition, and calculating the total distortion rate of the third harmonic; a graph attention network with 12-dimensional features is constructed, and weighted fusion is carried out through a multi-head attention mechanism; the lightweight CNN outputs a fault type, and when the confidence coefficient is less than 0.9, a knowledge graph rule engine is triggered; and updating a threshold value based on a historical diagnosis clustering result, and aggregating edge model parameters by federal learning. The system comprises a wafer-level micro-strain sensing layer, an FPGA accelerated edge computing layer, a cloud platform supporting federated learning, and an AR maintenance guidance and block chain evidence storage module. The early fault detection rate is improved, the false alarm rate under strong EMI is reduced, and the average repair time is shortened.

The invention relates to the field of artificial intelligence, and discloses a large language model illusion suppression method and system based on formalized proof, and the method comprises the steps: carrying out the logic mapping of the input and output of a language model through a formalized logic verification framework, and carrying out the logic verification of the output of the language model according to rule matching and recursive reasoning, identifying model illusion data; on the basis of an association rule mining algorithm, performing illusion mode mining on the model illusion data; clustering the illusion modes by utilizing clustering analysis to obtain a clustering result of the illusion modes; and based on a clustering result of the illusion mode, using a rule learning algorithm to formulate an inhibition rule, generating a branch rule of the decision tree, and performing illusion inhibition on an output result of the language model according to the inhibition rule and the branch rule. Through automatic rule learning and association rule mining, the occurrence of illusion modes is reduced, and the reasoning quality of the language model is improved.

The invention discloses an intelligent computing power cluster task allocation method and system based on cloud edge collaboration, and relates to the technical field of computing power task allocation, and the method comprises the steps: extracting a first task in to-be-allocated tasks, collecting and obtaining a first multi-dimensional feature parameter of the first task, and obtaining a first feature curve; performing clustering analysis on the to-be-allocated tasks to obtain a clustering result; judging whether distributed edge computing power in a cloud edge computing power cluster meets the first resource requirement or not; if yes, performing comparative analysis to obtain first fitness of the first clustering cluster; descending the computing power nodes to obtain a target computing power node; and performing task processing of the first clustering cluster. According to the method and the device, the technical problems of low task allocation efficiency and poor task and computing power node adaptability in the cloud-side collaborative environment in the prior art are solved, and the technical effects of realizing efficient allocation of the tasks in the intelligent computing power cluster in the cloud-side collaborative environment and improving the task processing adaptability and efficiency are achieved.

The invention provides a water-based adhesive coating control method and system based on artificial intelligence optimization, and the method comprises the steps: obtaining a historical data set composed of process parameters of a coating process and coating quality parameters, carrying out the clustering, and obtaining a global induction point set according to a clustering result; determining a stage induction point subset according to the target parameter value of the current coating stage and the boundary of the target parameter, and updating the sparse Gaussian process regression model by using the global induction point set and the stage induction point subset; calculating a quality fluctuation index, determining the length of a prediction time domain based on the quality fluctuation index, constructing an optimization problem in the prediction time domain by adopting a multi-level opportunity constraint mode, and when the deviation value between the actual value of any key process parameter and the prediction trajectory based on the sparse Gaussian process regression model exceeds a deviation threshold value, determining that the prediction trajectory does not exceed the deviation threshold value. And solving the optimization problem to obtain an optimal control action sequence in the prediction time domain, and determining a final control action from the optimal control action sequence and sending the final control action to an execution mechanism.

The invention relates to the technical field of artificial intelligence, in particular to a vehicle condition data exception handling system based on clustering analysis, which comprises a data vector module for processing according to collected historical normal sample data and calculating a first data vector; the vector processing module comprises a data acquisition unit, a running state vector unit and a judgment unit; the data acquisition unit constructs a multi-dynamic sensor network and acquires real-time data of the motorcycle in accordance with the first data vector; the operation state vector unit processes frequency spectrum information of the data in the real-time data through Fourier transform, and calculates a multi-feature information fusion vector; the judgment unit establishes a clustering analysis model to process the obtained multi-feature information fusion vector to obtain a first clustering result and a second clustering result; and the controller obtains a second clustering result and performs automatic adjustment according to a reinforcement learning algorithm.

The invention discloses an index optimization and compression storage system and method for a large-scale literature set, and the method comprises the following steps: S1, collecting and preprocessing literature data, and generating a standardized text data set; s2, carrying out keyword semantic vector coding, and constructing a keyword semantic vector matrix; s3, constructing an initial Gaussian mixture model to obtain a clustering center, a covariance matrix and a weight; s4, introducing a sea elephant optimization algorithm to optimize clustering parameters, and outputting an optimal clustering result; s5, constructing a semantic clustering structure, and generating an index tree structure; s6, performing bitmap compression and inverted coding, and constructing an index table supporting Boolean logic; and S7, dynamically accessing the newly added literature, and completing incremental updating of the index structure. The method is used for improving the index construction efficiency and the storage compression rate of a large-scale literature set, and efficient and semantic literature retrieval service capable of being incrementally updated is achieved.

The invention discloses a road broken line automatic repair method based on unmanned aerial vehicle image segmentation and fitting reconstruction, and relates to a remote sensing image semantic processing and road connectivity recovery method. The method comprises the steps of firstly extracting a road category region in an image; skeletonizing the extracted road area to form a road trunk structure; introducing a multi-scale density clustering algorithm to perform multi-level spatial clustering on the breakpoints, and superposing clustering results to enhance breakpoint coverage; screening out breakpoint pairs conforming to broken line characteristics; for each pair of matched breakpoints, backtracking along the road direction to obtain a certain number of skeleton points, and reconstructing a road connection path; and restoring the repaired road in combination with the width information of the original road to form a complete connected road structure. The method can effectively improve the automatic detection and continuity recovery capability of the road broken line area in the remote sensing image, and is suitable for a road structure intelligent restoration task under a low-quality image condition and a complex road network scene.

The invention relates to the technical field of intelligent driving, and particularly provides a personalized driving behavior identification method based on deep embedded clustering. Extracting a low-dimensional depth feature vector of the driving time sequence segment through a deep neural network auto-encoder, performing end-to-end joint optimization on the auto-encoder and unsupervised clustering by using a deep embedded clustering method, generating a driving style clustering result, and labeling a clustering result label; freezing the trained auto-encoder parameters, constructing a driving style classification model, and establishing a mapping relation between driving time sequence fragments and style labels; and distributing semantic tags for clustering results by extracting interpretable features. According to the end-to-end joint optimization technology based on deep embedded clustering, the convergence speed and generalization ability of the model are remarkably improved, and the driving style identification precision is far higher than that of a traditional machine learning method.

The embodiment of the invention provides a risk zoning method and device for helicopter flight, equipment and a storage medium. The method is applied to the technical field of data processing, and comprises the following steps: carrying out classification and threshold division on meteorological elements according to a fusion data set and different flight scenes to obtain a meteorological element threshold breakpoint set; based on the meteorological element threshold breakpoint set, performing frequency statistics and data synthesis on the fused data set to obtain a multi-layer meteorological element frequency data set; obtaining an initial clustering result of each meteorological element through clustering analysis according to the meteorological element frequency data set; according to the initial clustering result, carrying out risk grade division on each meteorological element to obtain a standardized single-element risk zoning result; and based on a single-element risk zoning result, adopting a multi-element fusion method to obtain a comprehensive risk zoning map. In this way, the technical problem that in the prior art, due to the fact that multiple elements jointly participate in clustering, the risk level is difficult to judge can be solved.

The invention discloses a multi-layer heterogeneous sequential network feature alignment and clustering method and device based on tensor self-representation. The method comprises the steps that sequential feature matrix sequences {},..., {} of data views in a multi-view heterogeneous network in time windows are obtained; based on the matrix sequence of the first time window, constructing a Laplacian regular term by introducing a self-representation learning mechanism, modeling multiple views by adopting a tensor structure and introducing a Schatten p-norm regularization mode to construct an optimization objective function for optimization, and representing shared feature representation of each data view in the first time window; and solving the function and carrying out clustering processing on the obtained function, wherein the clustering result of each time window is used for carrying out characteristic analysis on nodes in the network. According to the method, on the premise that the original structure and semantics of the network are guaranteed, unified modeling can be carried out on the incomplete multi-view heterogeneous network, and feature expression is aligned.

The invention provides a radar-based off-bed detection method, system and product, and the method comprises the steps: carrying out the signal preprocessing of a collected radar echo signal, and obtaining a multi-dimensional echo signal containing the information of a distance dimension, a speed dimension and an angle dimension; performing target detection based on the multi-dimensional echo signal, and generating a target 2D point cloud based on a target detection result; clustering processing is carried out based on the target 2D point cloud, feature extraction is carried out based on a clustering result, and a feature vector is generated; and inputting the feature vector into a pre-trained human body in-bed state judgment model to carry out off-bed event detection. According to the off-bed detection method provided by the invention, actual image data is not needed, real-time point cloud of nursing personnel on the bed can be provided under the condition of not invading privacy, the off-bed event can be reported in time, rapid detection of the off-bed event of long-term bedridden old people can be realized, the off-bed detection response time can be effectively shortened, and the detection efficiency is improved. And the accuracy and the real-time performance of off-bed event detection are effectively improved.

The invention discloses a knowledge data de-duplication method and device, a storage medium and computer equipment, relates to the technical field of data processing, is suitable for businesses such as financial science and technology and smart medical treatment, and mainly aims to solve the problem of poor de-duplication effect when large-scale knowledge data is processed in existing knowledge data de-duplication. Comprising the steps of obtaining large-scale knowledge data of related businesses; performing clustering processing on the large-scale knowledge data by adopting a MinHash LSH model to obtain a repeated text clustering result; the repeated text clustering result comprises a plurality of groups of similar data sets; performing word embedding calculation on each group of similar data sets by adopting a bge-m3 model to obtain word embedding corresponding to each group of similar data sets; and respectively carrying out semantic similarity de-duplication processing on the word embedding in each group of similar data sets to obtain a de-duplication result of the knowledge data.

One or more embodiments of the invention provide a user behavior trend clustering method and device. According to the method, a plurality of user behavior sequences corresponding to a target service are obtained; performing weighted calculation on a node characteristic value of each node in each user behavior sequence; wherein the node weights corresponding to the nodes of different behavior types are different; performing weighted calculation on a node relation characteristic value of each directed node pair in each user behavior sequence; wherein the directed node pairs of different path types correspond to different relation weights; according to a preset fusion weight, performing weighted fusion on the node feature value and the node relation feature value to obtain fusion features corresponding to the user behavior sequences respectively; and performing clustering processing based on each fusion feature to obtain a clustering result.