184 results about "Relationship mining" patented technology

The present invention relates to an entity relationship recognition method and apparatus. The method comprises obtaining a statement sequence from a target text in a corpus, and performing named entity recognition and dependency grammar marker on the statement sequence to obtain a marked text sentence; matching and retrieving the marked text sentence on basis of an entity relationship seed to obtain a training example; replacing the entity relationship seed word in the training example with predetermined identification, processing the training example after replacement combined with the named entity recognition and the dependency grammar marker, and generating a candidate rule; fuzzifying the candidate rule to obtain fuzzy rules; determining whether the fuzzy rules comprise a new rule; and retrieving the corpus according to the fuzzy rules to obtain a seed set when the fuzzy rules comprise the new rule, and using the obtained seed set as an entity relationship recognition result. Manual participation can be effectively reduced, dependence on the calibrated corpus is reduced, a new entity relationship can be found timely, and the entity relationship recognition method and apparatus are self-adaptive to entity relationship mining in different fields.

The invention relates to a method for constructing a knowledge graph based on entity extraction and relationship mining of a rule model. The method comprises the following steps: step 1: crawling data of an encyclopedia knowledge base of a target region, and defining dictionaries of foods, pesticides, nutrition and plant diseases and insect pests, so as to be convenient for rule mining; step 2: carrying out HTML (Hypertext Markup Language) label removal on encyclopedia type data to obtain Chinese texts and obtaining a URL (Uniform Resource Locator) link, so as to be convenient for subsequent processing; step 3: obtaining more complete entity attribute information by adding manually annotated relation attribute information; and step 4: obtaining an event and establishing a graph relation. According to the method provided by the invention, text information is converted into word vector mathematical information; vector similarity comparison is carried out and a relation between entities is labeled according to a relation between numbers, so as to represent a core knowledge base for the field and improve and optimize search quality; and a process from a simple character string to entity comprehending is realized.

Efficiently mining frequent subgraphs contained commonly in a plurality of objects by considering abstraction. A frequent pattern mining apparatus which mines a set of frequent subgraphs contained in a plurality of objects including: a conceptual hierarchy database which stores correspondence between multiple predetermined types of component and a superordinate conceptual element; a mining candidate subgraph generator which generates a set of mining candidate subgraphs using the frequent subgraphs which have already been mined; a weighted support calculator which calculates weighted support of the mining candidate subgraphs; a superordinate concept candidate selector which selects a second mining candidate subgraph to be a superordinate conceptual pattern of a first mining candidate subgraph; and a frequent subgraph adder which adds the second mining candidate subgraph to the set of mining candidate subgraphs if the first mining candidate subgraph and the second mining candidate subgraph are not identical in weighted support.

The invention provides a spam detection device and a method based on user relationship mining and credit evaluation. The detection device comprises a user relationship modeling device, a fingerprint collecting device, a path mining device, an email evaluating device, an email outputting device, a user evaluating device and a credit updating device. The method comprises the following steps: establishing a user relationship model according to the user communication relationship in a controlled network, then mining a specific path set according to the user relationship model, and finally establishing email determination records through the historical evaluation of the user on the path set and identifying email attributes. The device has the advantages that the device has good identification ability and can well identify malicious users and resist the interferences of the malicious users, can be deployed at entrances and exits of an email server, a gateway and a backbone network and be widely applied in the email filtering technology and other application fields.

The invention discloses an entity relationship mining method based on biomedical literature, which comprises the following steps: (1) querying disease-related biomedical literature in a public database, and obtaining biomedical text data after data preprocessing; (2) performing biomedical named entity recognition on the obtained biomedical text data in combination with a regular matching templateand a deep learning model; and (3) based on the entity recognition result, mining the entity relationship by adopting transfer learning and reinforcement learning methods. The biomedical nouns entities in the literature can be effectively identified by acquiring the biomedical literature related to the disease from the network, extracting the abstract and the title and carrying out entity recognition and relationship mining, and the hidden relationship among various entities can be mined.

The invention discloses a road traffic flow prediction method, and the method comprises the steps: sampling the historical traffic information of a road condition, and constructing a traffic network model; providing a missing data complementation method based on historical traffic information, building a reachable matrix according to travel time, carrying out the graph convolution through employing the reachable matrix as a convolution kernel, and extracting features; training a recurrent neural network according to the extracted features and the collected traffic data to obtain a flow prediction model; based on real-time sampling traffic data, inputting a pre-trained prediction model, obtaining a new error result in the process, and dynamically training the model.The reachable matrix canreduce unnecessary space search, the space-time relationship mining efficiency is improved, and the road traffic flow prediction accuracy under the conditions of more missing of road traffic flow dataand longer prediction duration is improved.

The invention discloses a word relationship mining method and a word relationship mining device, which belong to the fields of computers and the Internet. The method comprises the following steps of: acquiring candidate relationships between two entries, the frequency of the candidate relationships and the word frequency of the entries; acquiring the statistic of mutual information and the statistic of a log likelihood ratio according to the candidate relationships, the frequency and the word frequency; acquiring a credibility normalization value according to the statistic of the mutual information and the statistic of the log likelihood ratio; and performing sequencing according to the credibility normalization value, and outputting the candidate relationships meeting a preset threshold as word relationships. The device comprises a first acquisition module, a second acquisition module, a third acquisition module and an output module. By the scheme provided by the embodiment of the invention, the accuracy of the mined word relationships is improved, and the user experiences are improved.

The invention discloses a friend relationship mining method combining network topology characteristics and user behavior characteristics. The method comprises the following steps of (1) establishing a friend relationship network diagram and randomly selecting 90% of friend relationship connection edge data as a training set and the rest of 10% as a test set; (2) constructing two weighted undirected graphs of a friend relationship network based on topological similarity; (3) constructing two weighted undirected graphs of a friend relationship network based on user behavior characteristic similarity; and (4) utilizing a community detection algorithm (CNM algorithm) based on weighting modularity to respectively carry out community division on the four weighted undirected graphs, if any two users are classified as one community in a division process of the four communities for at least three or more times, considering that the two users are friends. The method introduces the topology characteristics and the behavior characteristics are introduced into the user friend relationship networks and mines whether the two users are the friends through community division.

Provided are a method and device for mining a social relationship based on financial data. The method for mining a social relationship based on financial data of the present invention comprises: acquiring financial transaction data of a client user, and determining a financial transaction network according to the financial transaction data; determining a network topology attribute of the client user and a non-network topology attribute of the client user according to the financial transaction network; and determining a social relationship corresponding to the client user according to a topology attribute and a non-network topology attribute of the financial transaction network. The embodiments of the present invention solve the problems in the prior art that the efficiency of judging the social relationship between client users is low, and the social relationship of client users is not totally discovered.

According to the user relationship mining method and system based on the mobile communication data. The method includes: excluding the non-mobile phone number by collecting the original data ticket information; extracting user pairs with family relations and user pairs without family relations from historical users, and calculating related indexes according to the call ticket data and the positioninformation data to construct a training data and test data set width table; and establishing a supervised classification model by adopting a lightgbm algorithm taking the decision tree as a base classifier. According to the method, the supervised machine learning method is adopted, the relation between the users is accurately identified, the family relation of the users in the home network is identified, the family relation of the users in different networks can be identified, the homing relation and the friend relation of the home network are identified in an auxiliary mode, and good accuracy is achieved on the identification result.

An enterprise relationship mining method belongs to the field of data mining, and comprises the following steps of defining the relationships , wherein the enterprise relationships comprise a legal person relationship, a shareholder relationship, a job relationship, a branch mechanism relationship, an external investment relationship and a competition relationship; acquiring the data, wherein theenterprise data comprises the business license information, the shareholder information, the employee information, the branch mechanism information and the business range labeling information; cleaning the data, checking the data consistency, and processing the invalid values and the missing values; multi-source data fusion, integrating all the information obtained by investigation and analysis, and performing unified evaluation on all the information; extracting the relationship. The enterprise relationship mining is a core for constructing an enterprise relationship graph, and the enterpriserelationship graph can display the enterprise relationships to the users in the form of structured graphs, so that the users can quickly understand and further explore the enterprise relationships. The enterprise social circles, the enterprise investment circles, the enterprise stock right structures, the enterprise actual controllers, the enterprise risk assessment and the like can be discoveredby mining the enterprise relations.

The invention discloses a monitoring data intelligent sampling method based on relevancy analysis, which includes the four key steps: time series data encoding, relevance relationship mining, state transition matrix calculation, and state prediction. According to the method, a monitoring cycle can be dynamically adjusted according to the prediction on future resource usage of a main unit, thereby reducing sampling frequency while resource usage varies stably, and increasing the sampling frequency while the resource usage varies sharply to save computing and storage resources. Compared with the prior art, the method has the advantages that the monitoring cycle can be enlarged and sampling frequency can be decreased while a machine runs stably; when the machine running fluctuates, it is required to decrease the monitoring cycle and increase sampling rate; more meaningful monitoring data are acquired, the quantity of gibberish to be collected is decreased effectively, spending major computing resources to on gibberish acquisition, computing and other processing is avoided, efficiency is improved, and high accuracy is maintained while gibberish collection is reduced.

The invention discloses a knowledge graph generation method and device, a relation mining method and device, equipment and a medium, and relates to the field of knowledge graphs. According to the specific implementation scheme, the method comprises the steps of obtaining an initial knowledge graph, wherein the initial knowledge graph comprises a plurality of nodes with a connection relationship; determining a plurality of target map elements meeting set attribute information requirements from the initial knowledge graph, the map elements comprising at least one of a target node and a target edge; and combining the plurality of target map elements to generate a knowledge graph. According to the embodiment, the nodes and the edges can express more attribute information, and the expression capability of the knowledge graph is improved; moreover, the graph structure can be reduced, and when the knowledge graph is applied to business scenes such as dynamic relationship mining, the analysisprocess can be simplified, the calculation amount can be reduced, and the calculation efficiency can be improved.

The invention belongs to the field of cloud computing, and provides an application store adaptor, which provides the best application for users, and enables the application being free from installation and upgrading, and even free from limitation of operating systems, so that the user experience is improved. In order to achieve the purposes, the application store adaptor adopts the technical scheme that the application store adaptor comprises five modules of OpenAPI capture and registration, OpenAPI semantic annotation and relation mining, OpenAPI packaging, shortcut release and automatic recommendation, and application mashup. The application store adaptor is mainly applied to web application store occasions.

The invention discloses a proactive caching method based on content popularity and movement rules of users. By introducing social relationship mining, users in a small cellular network are subjected to community division, and according to a Zipf law, content popularity distribution of each community is analyzed; by a frequent pattern mining algorithm, the movement rules of the users are mined, andaccording to a rule confidence degree, probability distribution that the user accesses each base station in a next time period is acquired; and according to content request distribution of the usersand probability distribution of accessing to each base station, under limited cache capacity of each base station, a cache hit ratio is maximized by optimizing a caching strategy. By adopting the proactive caching method disclosed by the invention, a return load can be reduced, and a user delay is reduced; and meanwhile, the method has the decomposition characteristic and is low in algorithm complexity.

The invention discloses a data blood relationship mining method and device. The method includes the steps: counting data distribution information of each row of fields in a plurality of data sheets; comparing the data distribution information of one row of fields in one data sheet with the data distribution information of one or more row of fields in other data sheets; determining whether one rowof fields in one data sheet and one or more row of fields in the other data sheets have data blood relationship or not according to comparison results. The data distribution information of the fieldsin the data sheets is compared, the data blood relationship among the fields of the data sheets is determined, the method is convenient, low in cost and easy to implement, dependent intermediate factors are fewer, and the method can be universally used for mining the data blood relationship of various databases.

The embodiment of the invention discloses a police affair management method and device, equipment and a storage medium. The method comprises the following steps: determining user identity behavior data according to at least one of field acquisition data, police affair data and internet data; Determining relation characteristics of the user according to the user identity behavior data; And according to the relationship characteristics of the user, carrying out association relationship mining on the user. According to the embodiment of the invention, a big data analysis platform is provided foreach police service department in a police service environment; the identity behavior data of all users under the big data can be planned as a whole; The association relationship of all users in the coverage range of the big data can be effectively mined, so that effective retrieval, analysis and processing basis is provided for each service requirement in the police service environment, and the management and processing efficiency of each police service department on the police service data in the police service environment is improved.

The invention discloses a causal relationship mining method based on deep learning. The method comprises the following steps: firstly, preprocessing data by using technologies such as missing data supplement, data normalization and independent hot coding; then, based on a Keras deep learning framework, carrying out univariate time sequence prediction on the target characteristics by using an LSTM;adjusting the model structure and a series of hyper-parameters to obtain an optimal model; recording the R2 score of the model on the test set; and then, predicting all the alternative features by using the model to obtain R2 scores of the alternative features on the test set, subtracting the two scores to obtain a Granger causal relationship score of the alternative features and the target feature, and obtaining a quantitative number for describing the Granger causal relationship between the alternative features and the target feature. The method is suitable for the problem of influence factor analysis of other time sequences. In conclusion, the Granger causal relationship mining method based on deep learning has the advantages of mining much data in various fields.

The invention discloses an information retrieval and potential relation mining method for bridge management and maintenance information, and the method comprises the steps: S1, defining sample data structure information which comprises sample data and the relation between the sample data; S2, constructing a bridge management and maintenance ontology knowledge base structure based on the sample data structure information, and importing the instance sample data into the bridge management and maintenance ontology knowledge base structure to generate a bridge management and maintenance ontology knowledge base; S3, establishing an inference rule based on the bridge management and maintenance ontology knowledge base; and S4, acquiring to-be-mined information, and retrieving the bridge managementand maintenance ontology knowledge base and / or generating potential relationship information between the to-be-mined information by utilizing a semantic logic reasoning machine based on the reasoningrule. According to the method, the reusability and interactivity of the BIM model on the bridge management and maintenance information are improved, and the intelligent management level of the bridgemanagement and maintenance field is greatly improved by adopting the semantic logic reasoning machine to explore the potential relationship information in the management and maintenance information.

The invention provides a method for determining a network fault. The method comprises the steps of obtaining to-be-processed work order data; performing formatting coding on the to-be-processed work order data according to a preset coding rule to generate to-be-processed words, wherein the to-be-processed words belong to a pre-obtained dictionary; determining a target word from the to-be-processedwords; determining a word vector corresponding to the to-be-processed word according to the dictionary; determining an association relationship between a first word set and the target word accordingto the word vector association relationship mining model and the word vector corresponding to the to-be-processed word, wherein the first word set is a word in the to-be-processed word set except thetarget word; and determining M candidate words from the first word set according to the association relationship, wherein the work order data corresponding to the M candidate words and the work orderdata corresponding to the target word are used for determining the fault of the abnormal network state. According to the method, the accuracy of the association relationship of the work order data canbe improved, so that the network fault determination efficiency is improved.

The invention provides a character relationship mining model training method, a character relationship mining method and a character relationship mining device. The training method comprises the following steps: acquiring a space-time knowledge graph; carrying out random sampling according to a positive sample of the space-time knowledge graph, generating a negative sample, and determining head entity initial embedding, relation initial embedding, tail entity initial embedding and time embedding of the positive sample and the negative sample; performing vector rotation on the initial embedding of the head entity and the initial embedding of the tail entity to obtain quaternion embedding of the head entity and quaternion embedding of the tail entity; respectively replacing the initial embedding of the head entity and the initial embedding of the tail entity with corresponding quaternion embedding of the head entity and quaternion embedding of the tail entity to obtain a processed positive sample and a processed negative sample; and iteratively training the character relationship mining model to convergence by adopting the processed positive sample and the processed negative sample. According to the technical scheme, the relationship between entities evolved along with time change can be mined, and the integrity of the knowledge graph is improved.

The invention provides a high-resolution remote sensing image weak supervision building extraction method. The method comprises the steps of class activation image iterative optimization based on an iterative adversarial rising strategy, class activation image inter-pixel relation mining, class boundary detection, building pseudo label generation, semantic segmentation network training and building area extraction. Compared with an existing image-level weak supervision semantic segmentation method, the method has the advantages that an iterative adversarial rising strategy is used, a better building class activation graph can be generated, full mining of class activation graph information is achieved under the condition that extra supervision information is not introduced, and the class equivalence relation between pixels is obtained. Besides, the segmentation performance of the network on boundary processing is further improved by using a gated convolutional layer, so that an object region can be effectively expanded and covered in the boundary, and a high-quality building pseudo tag is generated, thereby enabling the semantic segmentation model to generate a building region with high accuracy and complete segmentation boundary.

The invention relates to an abnormity detection method based on a log event graph and association relationship mining. The method comprises the following steps: collecting an original log of a system to obtain log events; segmenting the log events into different groups according to a set time span or a task number, wherein the log events in each group form a log event sequence according to the generated time; according to association relationship mining, mining system log events having association relationships with each abnormity, and removing log events irrelevant to the abnormities in the log event sequence; extracting a semantic vector of each log event as a feature vector of the log event; generating a bidirectional full-connection log event graph according to the log event sequence, updating a feature vector of each node by using a gating graph neural network, performing weighted summation on the updated feature vectors of all the nodes by using an attention network, calculating a global feature vector of the log event graph, finally performing classification detection through a full-connection network, and obtaining the normal or abnormal type of the system.

The invention discloses an intelligent question and answer method and a system based on a tourism knowledge graph. The method comprises the steps of obtaining a to-be-queried statement; performing named entity recognition on the to-be-queried statement to obtain an recognized entity; performing entity relationship mining on the to-be-queried statement to obtain a mined entity relationship; according to the entity and the entity relationship, constructing a Cypher statement of the to-be-queried statement; and based on the Cypher statement of the to-be-queried statement, searching the tourism knowledge graph to obtain a corresponding answer. Local tourists are helped to quickly and efficiently obtain tourism related information, and tourists are helped to plan travel routes.

The invention discloses an assembly-oriented intelligent quality tracing method, and belongs to the technical field of computer integrated manufacturing technologies and automation. The method comprises the following steps of with a product assembling process s a main line, constructing an assembly relationship network model covering all key production elements such as equipment, personnel, materials and suppliers; mining a model through a quality problem influence factor association relationship, extracting and quantifying the incidence relation of each production element in the assembly relation network model so thatthe full coverage of complex, dynamic, related and non-linear relations among influence factors of the product assembly quality problem is realized, and the problems that the assembly quality problem tracing method is single and the tracing effect is inaccurate and not timely are solved.

The invention relates to a user relationship mining method and device, electronic equipment and a computer readable storage medium, relates to the technical field of big data analysis and mining. Theinvention can be applied to an application scene of analyzing user behaviors to complete user relationship mining. The user relationship mining method comprises the steps of obtaining user relationship data, and determining a user label of a to-be-processed user in the user relationship data; wherein the to-be-processed user comprises a target user and a neighbor user of the target user; obtainingthe maximum grouping number; performing label updating operation on the user label of the target user according to the maximum grouping number and the user label of the neighbor user so as to determine a result label of the target user; and determining a user group of the target user according to the result label. The user relationship can be analyzed through different dimensions, namely, coarse-grained analysis taking the community as the object and fine-grained analysis taking the user as the object.

The present invention discloses a non-limiting relationship mining method based on dependency syntax and a pattern rule. The method comprises the following steps: step one, performing word segmentation and part-of-speech tagging; step two, performing entity identification; step three, performing dependency syntax analysis; step four, performing entity relationship mining; and step five, obtaininga relationship mining comprehensive result. According to the method disclosed by the present invention, semantic information extraction of large-scale unstructured or semi-structured texts can be realized, automatic extraction of an entity relationship can be realized, entity relationship information can be effectively mined from a large number of texts, and business application value can be enhanced.

The invention provides a directory mapping relationship mining method. The method includes the following steps: taking a full-amount entry under single-class classification in an entry system as an entry to be mapped and taking annotation data and a synonym list as a mapping dictionary; respectively performing user directory name mapping and direction content mapping to determine a standard directory preliminarily corresponding to a user directory name; determining a standard directory finally mapped by the user directory name by means of empowerment voting. Correspondingly, the invention further provides a directory mapping relationship mining device. By directory mapping relationship mining, overall readability of the entry system can be effectively improved.

The invention discloses a massive data analysis method based on a cloud platform. The massive data analysis method comprises the following steps: S1, obtaining biological data from a gene sequencing device, and storing the obtained biological data in a cloud centre server; S2, dynamically setting a keyword, and acquiring the biological data matched with the set keyword from the cloud centre server through a network; S3, classifying the acquired biological data, and storing the classified biological data in corresponding databases; S4, sequentially performing information analysis, field analysis and relationship mining analysis of the biological data stored in the databases so as to obtain an analysis result and an analysis process framework, and establishing a data index table according to the analysis result; and S5,displaying the analysis result and the analysis process framework by calling the obtained analysis result and the obtained analysis process framework, and uploading the analysis result and the analysis process framework to the cloud centre server. By means of the embodiment of the invention, the analysis method is high in efficiency, clear in analysis result, good in readability and convenient for users to share.