68 results about "Association mining" patented technology

The invention discloses a cross-media retrieval method. Cross-media retrieval can be carried out through characteristic mapping based on coupling and association mining in the same media or different media. The cross-media retrieval method comprises the following steps: establishing a multi-media data set; carrying out coupling characteristic mapping on all samples in a sample set; carrying out association mining in the same type of media; carrying out association mining in different types of media; obtaining a class probability matrix of images to be retrieved or texts to be retrieved through two association mining methods and weighting; respectively obtaining a class similarity matrix of image retrieval texts and a class similarity matrix of a text retrieval images, arraying similarities in a descending manner and retrieving. According to the technical scheme provided by the invention, the cross-media retrieval method is good in overall performance and good in applicability, and can be used in various cross-media retrieval scenes.

The invention provides an equipment test system planning decision-making method and system based on knowledge graph analysis, and the method comprises the steps: collecting test information in a tasktemplate mode, carrying out structured storage of the test information, obtaining original test data, carrying out information preprocessing of the original test data, and obtaining knowledge data; carrying out information extraction and knowledge fusion on the knowledge data by utilizing natural language processing, constructing a test knowledge graph, and presenting a visual test knowledge graph; and performing text mining on the test knowledge graph by adopting machine learning, and outputting a test system planning decision. According to the invention, test information is further organizedand abstracted into test knowledge, association between knowledge entity nodes and association mining between the knowledge entity nodes and metadata are carried out, finally, based on multi-attribute association of the knowledge nodes, a multi-source heterogeneous test knowledge graph is constructed, visualized presentation of the graph is carried out, and powerful support is provided for test system planning and decision making.

The invention provides a social network user interest association rule mining method based on community division. The method includes the steps that firstly, classification preprocessing is performed on data based on similarity, so that records in the same classification are high in similarity; secondly, association mining is performed on data of each classification to obtain frequent item sets of all the classifications; finally, all the frequent item sets of all the classifications are combined, and candidate association rules with the confidence coefficient larger than a confidence coefficient threshold value are selected to generate a final association rule collection. Generated unnecessary candidate item sets low in association performance can be reduced, so that the association rule mining efficiency of overall data is improved, and good expansibility is achieved.

The invention provides a method of forecasting electric quantity based on association mining of hot events, which has high practical value, reflects deep marketing rules and accurately forecasts the electric quantity. The method comprises the following steps: step I, analyzing data and building a model; step II, mining the hot events; and step III, mining the association among the events by using an association rule for electric quantity forecasting. The model is a feedforward neural network model, the hot events include space-time frequent events, exceptional events based on average line deviation and exceptional events based on the model, and the association rule refers to close correlation between itemsets occurring on a given training itemset frequently, wherein the 'frequency' is a threshold set artificially for measurement, namely support degree; and the 'close' is an association threshold set artificially for measurement, namely confidence level.

The invention discloses an enterprise credit risk early-warning and monitoring method, comprising: according to an enterprise's client list provided by the system, collecting multi-party data sources; cleaning the extracted data through the removal of invalid data and repetitive data in a manner of data combination and repeating the association mining; based on the obtained multiple data sources and in combination with the score evaluation card mode provided by related service people, determining the factors that meet the risk triggering condition and calculating the risk score of the enterprise itself; calculating the associated influence factor for the associated enterprise; setting different risk transfer weights and evaluating different transfer coefficients for the association relationship; and calculating the risk score of the associated enterprise; adding the calculated risk score weight of the associated enterprise to the risk score of a main enterprise; obtaining a final risk score; and finally, based on the calculated risk score, using a mechanics graph model to draw an association relationship graph. With the method of the invention, the presentation mode for the association relationship graph is optimized, and therefore, a better presentation effect can be achieved.

The invention discloses a power communication network alarm association mining method based on an improved GSP, and the method comprises the following steps: (1) collecting historical alarm data, building a historical alarm database, and carrying out the preprocessing of the historical alarm data; (2) generating a network topology model by analyzing the topological relation between the network element devices in the synchronous network topology database in the step (1); (3) performing data mining on the preprocessed historical alarm data in the step (1) by accessing the network topology modelassociation frequent sequence in the step (2) to form an alarm association rule and a sequence experience library; (4) collecting real-time alarm data, establishing a real-time alarm database, and preprocessing the real-time alarm data; and (5) performing data analysis on the preprocessed real-time alarm data in combination with an alarm association rule and a sequence experience library, and constructing an alarm real-time dependence model in combination with the network topology model in the step (2), thereby realizing positioning analysis of the root alarm.

The invention discloses a public opinion association analysis method based on a knowledge graph, and the method comprises the steps: extracting the attributes and relationships of entities in Internetpublic opinion knowledge, and constructing a public opinion business knowledge base based on the knowledge graph; determining a plurality of same or different types of entities needing association analysis, and combining the plurality of same or different types of entities in an intersection, combination or subtraction manner; and determining the result composition of association analysis of eachcombination mode of a plurality of same or different types of entities to obtain an analysis result. The invention further provides a public opinion association analysis system based on the knowledgegraph. According to the invention, association analysis of knowledge of the same type or different types including specific characters, specific organizations, specific events, specific topics and the like can be realized; and multi-dimensional deep analysis and association mining of the associated entities are realized, and the service users are helped to accurately grasp association conditionsof various different groups and intelligent analysis results of the associated entities so as to assist in decision making.

The invention discloses a scientific literature key content potential association mining method based on a graph neural network, and the method comprises the following steps: S1, obtaining scientificliterature data related to a certain specific event, and carrying out the data cleaning and preprocessing; s2, extracting a literature content keyword by utilizing a TF-IDF method; s3, constructing aword co-occurrence network for the extracted keywords and references to which the keywords belong by taking sentences as units; s4, learning vector representation of the keywords by using a graph convolutional neural network; and S5, obtaining the relevancy between different keywords by utilizing a similarity calculation function, and mining the potential incidence relation of the different keywords. According to the method, modeling is carried out on the keyword relationship extracted from the article content, and the potential association of the main keywords of the literature is mined by utilizing the graph convolutional neural network technology, so that the analysis requirement on the scientific literature content is met, and the correlation of the scientific literature in different fields is analyzed; and an effective method is provided for systematic analysis of scientific literatures.

The invention relates to a vulnerability association mining method for an industrial control system, and belongs to the technical field of network security. The method comprises the steps of: obtaining network vulnerabilities; classifying the vulnerabilities; adopting a machine learning algorithm and rule processing combined mode to automatically generating the whole permission set extraction process; and performing association analysis on the automatically acquired permission set fields contained in the description attributes, and mining a front-back logic relationship between industrial control system vulnerabilities. When the logic relationship between the industrial control vulnerabilities is analyzed, comprehensive description is carried out by using multi-dimensional attributes, theindustrial control vulnerabilities are effectively classified, besides basic description indexes, the concept of an authority set is also used, and the vulnerabilities are associated by using the privileged property of vulnerability attacks. The invention further provides a mode of combining a machine learning algorithm with rule processing, the whole process of extracting the permission set labelis automatically generated, and direct or indirect attack threats to an industrial control system can be caused by comprehensively and efficiently analyzing and utilizing security holes.

The invention relates to an automatic loophole classification method based on word frequency statistics and a simple Bayesian fusion model, and belongs to the technical field of information security. The method specifically comprises the steps of S1, building a loophole database, and collecting loophole records; S2, determining the class of a privilege set; S3, training a word frequency-reverse file frequency classifier; S4, testing a classification result and an accuracy rate of the word frequency-reverse file frequency classifier; S5, building a simple Bayesian classifier; S6, testing the classification result and the accuracy rate of the simple Bayesian classifier; And S6, fusing the classifiers. Compared with the existing methods, the automatic loophole classification method supporting loophole association mining provided by the invention has the advantage that a loophole description field in the loophole database is used, meanwhile, influence of loophole attributes such as availability scoring and influence scoring on loophole association are simultaneously considered, and classification accuracy is greatly improved.

The embodiment of the invention discloses a fault information processing method, device and equipment and a storage medium. The method comprises the following steps: acquiring service private line data of each service private line; performing association analysis on the multi-dimensional general fault data in the business special line data of each business special line by adopting a big data association mining mode, calculating a basic maintenance coefficient of each business special line in a preset period; analyzing multi-dimensional customer data in the business special line data of each business special line by adopting an artificial neural network, calculating a customer behavior maintenance coefficient of each business special line in the preset period; and performing binary linear regression analysis on the basic maintenance coefficient and the customer behavior maintenance coefficient to obtain the maintenance urgent degree of each business special line. According to the embodiment of the invention, the customer perception can be practically improved, and the customer satisfaction of an operator service special line is effectively improved.

The invention discloses a method and system for mining and searching unstructured text data in a financial field and provides a scheme of named entity identification for the unstructured text data, association relationship mining between entities and association relationship network construction and utilization in the financial field, and the scheme is mainly for information extraction in the financial field, named entity identification, association relationship mining and the construction and utilization of an association network. The system comprises a data acquisition module, a data cleaning module, a data preprocessing module, an association mining module, an association map construction module and a complex network analysis module. The invention can complete the basic data analysis and information extraction work, construct an economic map by using the mined information and use the economic map to mine deep information and a hidden association, so that the data have more intuitivereadability and utilization value.

The invention relates to a cross-media Hash method and system based on serialized multi-feature guidance. The method comprises the following steps of 1 establishing a database of images and texts, andrespectively extracting features of the images and the texts under various scales; 2 inputting different scale characteristics of the image and the text into two paths of recurrent neural networks according to a set sequence, and calculating Hash codes of the image and the text; and 3 optimizing network parameters through the inter-scale association constraint function and the inter-media and intra-media Hash constraint functions to realize inter-scale association mining and Hash function learning, and 4 in the retrieval stage, extracting different scale characteristics of the query image ortext, and generating corresponding Hash codes according to the same method in the step 2 to realize cross-media Hash retrieval. According to the method, the association relationships among various scales of different media can be mined to realize learning of the Hash function, and the higher retrieval accuracy is achieved compared with an existing method.

The invention discloses a regulation and control cloud-based power grid operation data association analysis method and system, and the method comprises the following steps: (1) employing an association rule classification and mining algorithm, and carrying out the data association analysis of the mass operation data of a power grid; and (2) establishing a relevance model of data and equipment states, and providing technical support for regulation and control cloud power grid operation risk pre-judgment and evaluation analysis. According to the invention, frequent item set mining is carried out by adopting a correlation analysis algorithm, correlation analysis is carried out based on power grid operation event data and historical data in regulation and control cloud general data object structured design, and an equipment defect and fault probability evaluation model is established; substation equipment operation state trend quantitative analysis is provided based on mass operation data of a power grid operation data platform, a behavior identification model of typical fault equipment is established in combination with alarm information, online monitoring data and equipment family information, and operation state and fault association mining and prediction analysis of power grid regulation and control equipment are realized.

The invention provides a multi-association mining method and device based on a social network. The method includes the steps of: for virtual humans having a known Internet account number, utilizing an association algorithm to mine other Internet account numbers of the virtual humans in the Internet; through a multi-identity association algorithm, processing all the Internet account numbers of the virtual humans, thereby obtaining a first-degree association result set of each Internet account number of the virtual humans; through a deep relation analysis algorithm, processing each first-degree association result set, thereby obtaining one or more important contacts of each first-degree association result set and an important contact account number set of each important contact; and for the important contact account number set of each important contact, utilizing a multi-association weighting algorithm to obtain a multi-association weighted value of each important contact. The method provided by the invention makes a quantitative analysis of connection between the virtual humans in the Internet, thereby showing the closeness of the connection between the virtual humans.

A targeted advertising system performs context-based association mining using a publicly available corpus to identify a product or brand name that, under a given context, is associated with a product or brand being marketed. The system analyzes documents within the publicly available corpus that are associated with the given context, and identifies products or brand names that have a high association to the product or brand being marketed. The system can also analyze the publicly available corpus to determine contextual information which is correlated to two or more products or brand names. This contextual information includes a set of terms that facilitates filtering the publicly available corpus into an optimal set of documents that has a high association to a desired market category or demographic.

The invention discloses a network data exception detection method and system based on high-order association mining, and the method comprises the steps: 1, generating a discrete forest according to anobtained network data set, and calculating a discrete value of network data in the network data set; 2, generating a clustering center point according to a clustering algorithm and the abnormal network data, and calculating a similarity value between the network data and the clustering center point; 3, calculating the weight of the network data in the network data set according to the discrete value and the similarity value; 4, calculating a label matrix of unmarked network data in the network data set according to the weight of the network data and the hypergraph model; And step 5, determining the security type of the unlabeled network data according to the label matrix. Through the technical scheme provided by the invention, effective identification of a large amount of unmarked networkdata in an industrial network environment is realized, and the identification accuracy of the unmarked network data and the utilization rate of the marked network data are improved.

A shelf layout optimization method for commodity data analysis of an unattended store specifically relates to knowledge in the fields such as electronic label commodity real-time data acquisition, association mining algorithms and frequent item set mining. The method is particularly applicable to analysis and processing of real-time data of commodities of an unattended store. An electronic label is used to collect commodity data; an association analysis algorithm is adopted, and performance of the algorithm is improved as much as possible from the perspective of optimizing a frequent item set;and by use of an improved degree-of-interest algorithm, an association among commodities is accurately found, so as to instruct optimization of a shelf layout of the unattended store. On the basis ofexisting commodity data analysis, the present invention provides a Top-K frequent item set mining algorithm based on a node set; and a POC-Tree data structure is used to compress transmitted data, sothat the limitation that the algorithm scans the database for multiple times is avoided, and the difficulty that the degree of support is hard to be set during the association rule mining process issuccessfully solved by means of better performance.

The invention relates to the technical field of production equipment maintenance, in particular to an association mining method and system for an abnormal event and a context event of the abnormal event. The method comprises the following steps: S1, acquiring an abnormal event table and a context event table; s2, obtaining a to-be-mined transaction table according to the abnormal event table and the context event table; s3, obtaining a mining result according to the to-be-mined transaction table, wherein the mining result is at least one of a context event most related to the abnormal event ora fault combination occurring at high frequency or a context event combination occurring at high frequency; and S4, displaying a mining result. According to the method, the context event most relatedto the abnormal event or the fault combination occurring at high frequency or the context event combination occurring at high frequency can be mined through the input abnormal event table and the context event table after processing, the method becomes an auxiliary tool for fault diagnosis and troubleshooting, and the method has the advantages of being easy to operate and low in cost.

The invention discloses a subway construction surface subsidence early warning method based on case-based reasoning and system dynamics. According to the adopted technical scheme, the method comprisesthe steps of constructing a case library based on domain knowledge; performing similarity retrieval based on case reasoning; risk association mining of case-based reasoning and establishment of a system dynamics model are established; and establishing a risk early warning model of the surface subsidence. The method has the beneficial effects that the reliable risk decision basis is obtained by optimizing the search case library, and the effective prediction and early warning can be carried out on the subway construction ground surface settlement risk.