443 results about "Rule mining" patented technology

A dense data-set mining system and method is provided that directly exploits all user-specified constraints including minimum support, minimum confidence, and a new constraint, known as minimum gap, which prunes any rule having conditions that do not contribute to its predictive accuracy. The method maintains efficiency even at low supports on data that is dense in the sense that many items appear with high frequency (e.g. relational data).

The invention provides a method and a device for fault locating, and relates to the application field of computer networks. The method and the device for fault locating solves the problem that an existing alarm relevance rule mining system is poor in timeliness and low in efficiency. The method includes constructing a network element topology constraint model; detecting running status of each network element device in a managed network so as to find fault events; collecting the fault events; and using the network element topology constraint model for time layer correlations and space layer correlations of the collected fault events so as to determine a fault location. The method and the device for fault locating are appropriate for fault diagnosis, and achieve efficient and accurate fault locating.

The embodiment of the invention discloses an alarm association rule mining method, an alarm association rule mining engine and an alarm association rule mining system. The method comprises the following steps of: acquiring an alarm sequence, wherein the alarm sequence comprises multiple pieces of alarm; calculating support of each k-item set to obtain a k-item frequency item set aggregate; generating a k+1-item frequency item set aggregate from the k-item frequency item set aggregate; calculating maximum degree of confidence of the k+1-item frequency item set aggregate according to the maximum value in the support of k+1 1-item sets included in the k+1-item frequency item set and the support of the k+1-item frequency item set for each k+1-item frequency item set in the k+1-item frequency item set aggregate; adding the k+1-item frequency item set with the maximum degree of confidence no less than minimum degree of confidence into an association rule result set as an association rule; and thus a false rule produced due to the influence of the parameter of the degree of confidence is reduced and the false rule in the association rule result set is effectively reduced during the mining of the alarm association rule.

The disclosure relates to technology for mining data in a database by recursively mining a conditional frequent pattern tree (FP-tree) for frequent items of each conditional pattern base for each node in an FP-tree to obtain frequent patterns. For each branch in the FP-tree, a single-item node table (NT) is generated for which a selected one of the frequent items appears in the node of the branch. The single-item NT including a list of all of the frequent items appearing in the FP-tree and a corresponding frequent item count. For each single-item NT of each branch generated for the selected one of the frequent items, the frequent item count of each frequent item is summed in the single-item NT formed for each branch to generate a combined single-item NT, and association rules based on the frequent patterns are generated for each of the frequent items and the combined single-item NT.

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.

A method and apparatus for generating computer configuration rules comprising receiving configuration data regarding a plurality of computers, analyzing the configuration data to determine associations within the configuration data, and generating configuration rules from a result of the analysis.

The invention discloses an association rule mining method based on mass data and relates to the technical field of information processing. The association rule mining method specifically comprises the steps that input data are divided according to records, a frequent 1-itemset table is set, data projection is carried out on the frequent 1-itemset table, the projected data are stored in a set grouped table, a local frequent pattern tree is set, then data association rule mining is carried out on the frequent pattern tree, in the process of mining on the frequent pattern tree, a pruning strategy is adopted, the number of iterations of the association rule mining is reduced, each computation node only processes one part of data records, and therefore the problem that the mass data can not be read into a memory by one computer to be processed is solved. In addition, various nodes participate processing in parallel, and processing efficiency is improved effectively.

The invention relates to an EHV (Extra-High Voltage) power grid fault rule mining method based on a rough set association rule. According to the method, a distributed data mining idea is adopted, historical fault data of an EHV power grid is subjected to fault rule mining by using an association rule mining method based on a rough set theory, a distributed decision table is subjected to attribute reduction by using an attribute reduction algorithm based on an information entropy, and then, an Apriori algorithm in association rules is applied to the decision table, which is subjected to reduction, so as to carry out fault rule extraction. According to the method, the problem of inadaptability to large-data-volume historical fault databases of traditional data mining methods can be solved effectively, the complexity of rule extraction is lowered, and the method has the advantage of high fault rule mining efficiency.

The invention relates to a power distribution network fault diagnosis method utilizing historical fault data. The method comprises the following steps: firstly, a fault information database is established from power distribution network fault rescue records, fault attributes contained in the fault information database are determined; then, the data format of the fault attributes is made to conform to the standard, and fault attribute data in the fault information database are discretized; an association rule mining method is used for mining a strong association rule contained in the fault attribute data in the fault information database; finally, according to actual conditions of a fault and the mined strong association rule, a diagnosis result of the power distribution network fault is obtained. The power distribution network fault diagnosis method utilizing the historical fault data is beneficial for improving safe operation of a power distribution network, and is high in reliability; besides, the power distribution network fault diagnosis method has the advantages of being wide in application rage, flexible to apply, and capable of being operated off line, and being influenced by the distribution network automation degree to a low extent, and the like, thereby providing a good basis for power distribution network fault diagnosis and state evaluation.

The invention discloses a resource individuation recommendation method based on user relevance. The method comprises firstly using a user relevance rule mining technique for analyzing history grading records of a user on resources and excavating a frequent set of a target user; then selecting one target user frequent set which is maximum in number of terms and highest in support to build an interest similar group of the target user; inputting history grades of the user in the interest similar group of the target user on the resources in a Slope One algorithm to serve as core data, and conducting grade forecast on resources without visiting of the target user; and finally recommending the resources without visiting and with a grade predicted value larger than the threshold value of the target user to the target user according to the value. Users with similar interests of the target user are used for forecasting in a process of grade forecast of the Slope One algorithm on the resources without visiting of the target user, grade matrix dimensionality of the target user and intermediately calculated data quantity are reduced, and accuracy rate of the grade forecast is improved.

The invention discloses a behavior characteristics-based network attack detection method and device, and relates to the technical field of information security. The technical points of the method comprise the following steps: step 1, collecting original security information output by various types of security equipment, and converting all security information into security events with the unified format; step 2, classifying all security events according to each field content; and step 3, sequencing all security events which have the same source IP address and the same destination IP address and occur within one monitoring period according to the sequence of event generation time to obtain security event combinations, and searching whether the same security event combination is stored in a security event correlation rule library or not, if yes, determining that a host corresponding to the destination IP address suffers from attacking and warning, or if not, storing all security events in an association rule mining database.

The invention provides an association rule mining method of large-scale data, and the method comprises the following steps that (1) the input data is subjected to classified preprocessing based on similarity, so that records in the same category have high similarity; (2) the data in each category is mined based on Apriori algorithm to obtain frequent item sets of all categories; and (3) the frequent item sets of all the categories are merged, and association rules which correspond to the frequent item sets which are more than the minimum confidence coefficient are determined to be strong association rules. According to the association rule mining method of large-scale data, unnecessary candidate item sets with small association can be reduced, so that the association rule mining efficiency of all the data is improved, and better expandability is realized.

The invention discloses an association rule mining method and a system thereof. The method comprises the steps of: generating a K+1 item set from a frequent K item set; performing a plurality of parallel processing tasks, wherein each processing task obtains data of the corresponding part in a transaction data set, and counting the frequent count value of the K+1 item set in the data; collecting the statistical result of all processing tasks to obtain the frequent count value of the K+1 item set in the transaction data set; generating the frequent K+1 item set which meets the requirement of support degree according to the frequent count value of the K+1 item set; and outputting the association rule when the association rule which meets the requirement of support degree is judged to be existed according to the frequent K+1 item set. The invention can improve the processing efficiency for mining the association rule.

The invention discloses a method and a device for realizing an association rule mining algorithm supporting a distributed computation. An HDFS (Hadoop Distributed File System) programming model is used to carry out two-stage analysis of a map function stage and a reduce function stage on the association rule mining algorithm, and the analysis steps comprises the following steps: step 1, a job scheduler is configured; step 2, a data set is read by a prior probability mapping module, and the data of the data set are converted by a map function into a value pair; step 3, the value pair processed in the step 2 is read by the prior probability reduction module, an ordering rule Top N containing an i item set is randomly generated by a reduce function, and the prior probability distribution value of a confidence coefficient is calculated at the same time; step 4, the same data set is read by a rule mapping module, and the data row of the data set is converted by the map function into the value pair; and step 5, the value pair processed in the step 4 and the prior probability distribution value in the step 3 are read by a rule reduction module, and the predication accuracy value of the ordering rule Top N is calculated by the reduce function. The method and the device for realizing the association rule mining algorithm supporting the distributed computation are mainly applied to the PA (Pridictive Apriori)-distribution type computing technology.

The invention provides a thermal power plant main operating parameter target value determining method based on association rule mining. The rule knowledge of a unit under certain operating conditions is acquired by using steady operating data stored in a historical operating database through an association rule mining technology, and finally the knowledge is used to guide the unit optimization operating. The method comprises the steps that the steady operating data are extracted from a unit historical operating data set through a data preprocessing technology; discretization of continuous attributes is carried out on a specific condition data set through a condition dividing result to complete data set preparation of association rule mining; and rule knowledge extraction is carried out on a discretized data set through an association rule mining technology to acquire a unit historical operating knowledge database. The knowledge database can assist an operator or a control system to make relevant decisions. According to different target attributes, the economy, stability and environment protection of unit operating can be improved, and a unit is under an excellent operating condition for a long time.

The invention provides a method for building an alarm association rule mining system based on a maximal frequent item set DM (Data Mining) and the realization of that. Three different mining ways for single equipment, cognate equipment and linked equipment are designed aiming to different types of equipment in a communication network, and according to the mining range, the DM range can be positioned at specific city level or communication equipment manufacturer level; after the mining way is confirmed, the alarm association time window, the sliding step, etc are selected to acquire an alarm affair set assembly; and alarm association result mining is performed by utilizing the maximal frequent item set mining algorism after a user inputs the minimum support, and mining result treatment and display are performed according to different mining ways. Through the mentioned steps, the alarm association rule needed by the user can be found out from plenty of alarm data. The method has broad application prospect and favorable utility value.

The invention provides a multi-dimension and multi-level association rule based voltage sag predicting and analyzing method, belonging to the technical field of power quality analysis methods. The method comprises the steps of selecting the sag association rule mining dimension; discretizing historical data; mining a sag association rule according to the minimum support and minimum confidence degree; building a voltage sag association rule knowledge base; and matching the association rule to obtain the prediction conclusion. The method has the beneficial effects that the historical voltage sag association rules are mined, the strong association rules obtained after mining form the knowledge base, and the grid operation conditions probably occurring in the future serve as the prediction conditions and are input into the rule base to be matched, so that the voltage sag conditions probably occurring in the future can be obtained. The method is a major supplement to the existing intelligent power quality monitoring system, and has very great practical significance.

The embodiment of the invention discloses an alarm correlation analysis method, device, equipment and medium. The method comprises the following steps of: acquiring alarm data, and carrying out line segment deduplication and standardization on the alarm data to obtain standardized alarm data; corresponding each alarm data according to the standardized alarm data and the characteristic field; excavating an alarm association rule according to a preset alarm rule mining parameter based on an alarm family spectrum so as to obtain an alarm association rule after mining; obtaining a preset alarm social network according to the mining alarm association rules; presetting to carry out alarm analysis on the real time alarm data, according to the alarm family spectrum, the minded alarm association rule, and the alarm social network. The invention can effectively de-emphasize the massive alarm data, improve the calculation efficiency through association rule mining, and quickly grasp the main features of the alarm storm, and the complex relationship between the alarms, in particular to the root cause alarm the chain alarm occurs. Derived alarms have a more in-depth analysis.

The invention provides an association rule mining method for privacy protection under a distributed environment. The association rule mining method is used to carry out global mining on multiple data and comprises the steps of: structuring a random disturbance matrix of item sets, carrying out disturbance transformation on data, making statistics on the summation of supporting number matrixes after disturbance, restructuring data distribution, precisely calculating the global support degree of the item sets in a space after pruning, and the like. According to the method disclosed by the invention, by means of structuring the random disturbance matrix to disturb a plurality of attributes at the same time and taking the correlation among the attributes into consideration in a disturbance process, the recover precision is effectively improved; after the supporting number of the item sets is evaluated by using a disturbance method, the final global frequent item set is determined by secure multi-party computation after pruning is carried out based on minimum support degree, thus, the communication traffic is effectively reduced, the mining efficiency is improved, a better compromise between the mining efficiency and the mining precision can be acquired, and the association rule mining method has a wider application range.

The invention provides a large-scale mixed heterogeneous storage system-oriented node fault prediction system and method. A time sequence-based association rule mining algorithm is adopted to construct a node fault prediction system architecture, and a main process of node fault prediction includes: collecting state data and log information of each storage node; carrying out data preprocessing, and generating sequence modes on the basis of a sliding window; using the sequence modes and fault sequences, which are extracted in a fault identification process, together as input of an association rule algorithm, and outputting output results as typical fault sequences; carrying out matching on the typical fault sequences and sequence modes generated in real time; and if a matching result meetsan established rule, issuing early warning to notify a system administrator, and giving feedback to a prediction result by the administrator according to a subjective interest degree. According to thesystem and method, real-time online fault prediction is carried out for nodes of a large-scale mixed heterogeneous storage system, and accuracy and recall which are better than those of existing fault prediction algorithms and better scalability can be obtained.

The invention relates to an abnormal behavior detection method and system. The method comprises steps as follows: acquiring to-be-detected behavior information of a user, and calculating the matching degree between the to-be-detected behavior information and historical abnormal behavior information; screening out the historical abnormal behavior information with the matching degree higher than a first preset threshold; acquiring an abnormal behavior sequence corresponding to the screened-out historical abnormal behavior information, and acquiring an association relationship between the screened-out historical abnormal behavior information and association behavior information corresponding to the historical abnormal behavior information in the abnormal behavior sequence; acquiring the association behavior information of the to-be-detected behavior information according to the association relationship, and constituting a to-be-detected behavior sequence by the to-be-detected behavior information and the association behavior information corresponding to the to-be-detected behavior information; calculating similarity of the to-be-detected behavior sequence and the abnormal behavior sequence; acquiring to-be-detected behavior information with the similarity higher than a second preset threshold, and determining the acquired to-be-detected behavior information as abnormal behavior information. According to the abnormal behavior detection method and system, abnormal behavior detection for the user can be performed accurately.

The invention discloses a power-network-dispatching-oriented intelligent warning analysis method. The method comprises the following steps that warning information of a power network is extracted and is preprocessed; denoising is carried out on the preprocessed warning information, and a denoising classification rule is obtained; the denoised warning information is subjected to induction and dupliction removal, warning combinations are generated, and the warning type to which each warning combination belongs is judged; each warning combination and the corresponding warning type are subjected to association rule mining, and a power network warning inference rule is obtained. After the intelligent warning analysis method is utilized, the filtering rate of noise data in the warning information and the comprehensiveness of the warning types are improved.

The invention discloses a boiler control method and device based on association rule mining. The method includes the steps that state data of a boiler are collected; preprocessing is performed on the data, piecewise linearization approximation is performed on the data, the data are converted into vectors in a two-dimensional space and then subjected to time series fitting, clustering is performed on all time series, and then time series flows are converted into transaction flows; a global SWFI-tree and a local SWFI-tree are established, information of the local SWFI-tree is added to the global SWFI-tree, and the global SWFI-tree is pruned; frequent item sets are generated according to an FP-growth algorithm; by the utilization of preset confidence coefficients, the association rules are generated through the frequent item sets; the association rules are used for predicating change states and trends of all parameters of the boiler after appointed time; according to predication results, the parameters are modified in advance, and the boiler is controlled to operate. Uniform mining is performed on the state data of the boiler, and the mined association rules are used for modifying the parameters of the boiler, so that the purpose of intelligently controlling the boiler to operate is achieved.

The invention discloses an industrial big data driven total completion time prediction method and relates to the field of engineering application. The method includes: establishing an industrial big data analysis platform; applying an association rule mining algorithm to analyze and mine total completion time influence factors; establishing a neural network model BP; dynamically improving a weight and a threshold of the neural network model BP to acquire a dynamical neural network model DBP; applying an AIGA (adaptive immune genetic algorithm) to optimize the dynamical neural network model DBP so as to acquire a prediction model AIGA-DBP, and computing a total completion time prediction value according to the prediction model AIGA-DBP; when an error of the total completion time prediction value and a total completion time expectation value meets preset conditions, outputting the total completion time prediction value. By the method, the total completion time can be predicated accurately, the work flow of enterprises is optimized, the production efficiency of the enterprises can be improved, and the method is adaptable to various changes of the enterprise due to time lapse.

The invention discloses a method for constructing information system running rule libraries on the basis of association rule mining. The method is characterized by comprising steps of S01, acquiring network topology architectures of information systems and dynamic monitoring indicators and static monitoring indicators of all devices; S02, generating network fault trees by the aid of the network topology architectures and the dynamic monitoring indicators and the static monitoring indicators of the devices and generating basic rule libraries by the aid of the network fault trees; S03, executing association rule mining algorithms on historical data of the information systems to acquire association rule libraries; S04, combining the basic rule libraries with the association rule libraries and generating extension rule libraries by means of reasoning. Retrieval priority of the basic rule libraries is superior to retrieval priority of the association rule libraries, and the retrieval priority of the association rule libraries is superior to retrieval priority of the extension rule libraries. The method has the advantages that the information system running rule libraries can be intelligently generated by the aid of fault tree technologies and association rule mining technologies, rules can be optimized by the aid of machine learning technologies, three-domain structures of the rules are further designed, and accordingly the rules can be automatically sorted and adjusted.

The disclosure relates to the use of one or more association rule mining algorithms to mine data sets containing features created from at least one plant or animal-based molecular genetic marker, find association rules and utilize features created from these association rules for classification or prediction.

The embodiment of the invention discloses a method and a device for data mining of a road traffic accident based on an association rule, which relate to the field of communication. In order to realize effective analysis and counting correlative with occurring condition factors of the accident and serve the jobs of accident prevention, treatment, decision and the like, the invention provides a technical scheme of: acquiring historical data in a prescribed time range and a prescribed road section range from the historical data of the road traffic accident as association rule mined data; determining the attribute of the traffic accident of the association rule mined data; acquiring a frequent item set according to the association rule mined data, the attribute of the traffic accident and the minimum support; and acquiring an accident association rule according to the frequent item set and the minimum degree of confidence. The method and the device are suitable for real-time dynamic traffic information services.

The invention relates to a method for analyzing traffic congestion reasons and particularly relates to a method for judging traffic congestion reasons based on characteristic significance. The method is characterized by extracting traffic congestion levels and relevant characteristics of congestion point sections, obtaining a reason library of the congestion point sections based on a prediction residual integration characteristic importance assessment method and analyzing specific reasons of one congestion based on a association rule mining method; assessing characteristic importance by using multiple supervised learning methods including an Lasso model method, a random forest model and a linear model method, and weighting the characteristic importance assessed by three methods according to prediction errors. Through the characteristic importance analysis, more accurate and more robust results can be obtained, so that the accuracy of analysis of the traffic congestion reasons is ensured.

The embodiment of the invention discloses an abnormal behavior detection method and device, relates to the technical field of network security, and can perform association rule analysis and similaritymining on audit data of a user so as to timely discover abnormal behaviors of the user. The method comprises the following steps: carrying out data type conversion on acquired historical audit data to generate first data; Wherein the auditing data comprises at least one operation record of accessing the network by the user through the client, and the operation record comprises but is not limitedto one or more of the following items: a timestamp, a client IP address, a target network IP address and an operation type; Calculating the first data according to an association rule mining algorithmto generate a historical user association sequence; And performing similarity mining on the historical user association sequence and the acquired real-time audit data to generate a similarity score,and comparing the similarity score with a preset threshold value to determine whether the user operation has an abnormal behavior or not. The embodiment of the invention is applied to a network system.

The invention discloses a knitting MES production planning and scheduling method based on big data mining, wherein the method belongs to the field of textile engineering application. The method comprises the following steps of S1, establishing a multidirectional knitting production data model; S2, establishing a big data analyzing platform based on a Hadoop distributed platform; S3, mining a planning scheduling restraining factor in a MapReduce frame by means of an Apriori correlation rule mining algorithm; S4, analyzing an order arranging priority according to a contract requirement; S5, integrating the arranging priority and the planning scheduling restraining factor, and obtaining an order arranging Gantt chart; and S6, performing real-time monitoring on ERP and ZigBee data, and when abnormities such as order delivery data change, knitting process updating and knitting machine sudden fault occur, dynamically adjusting a production plan.