Fuzzy Association Rule Mining Based On CHC Algorithm

With the application of "Internet+" technology and the rise of emerging industries of Big Data,a large amount of data is accumulated in the databases of various fields,and this vast amount of data hides many valuable information.Association rule mining can obtain potential and valuable associations in a huge database.However,quantified browsing frequency can not be a good measure of user's interest in mining association rules for web page.And the NFDMA algorithm is proposed.In this thesis,using fuzzy language variable to describe the user's browsing web pages.NFDMA algorithm is studied in the following three aspects:Firstly,the 3-tuples linguistic model represents the membership function.Based on the two-tuples,this model considers the fuzzy region.The dependency among membership functions and the dependency among the three definition points,lead to tuning models handling very complex search spaces which affect the good performance of the optimization methods.Secondly,The CHC algorithm(cross-generation heterogeneous recombination large variation)optimizes the membership function of the triple semantic model representation.According to the given membership function,the browsing frequency is transformed into a fuzzy set,but the membership function given by the random is not accurate,so this thesis uses the CHC algorithm to optimize the membership function of the triple semantic model representation.Thirdly,build the FP-tree to mine fuzzy association rules.Using FP-growth algorithm make the optimal membership function obtained in NFDMA algorithm to mine fuzzy association rules by constructing tree structure.In the traditional way,the association rules are represented as:If A Then B.and the form of the 3-tuples linguistic representation model is:If A(?1,?1,?1)Then B(?2,?2,?2).The variable?1,?2 represent the fuzzy linguistic variables,indicating the level of interest of people.The variable ? and ?,representing the lateral displacement and the amplitude variation of a fuzzy area.The fuzzy association rules derived from the triple semantic model can not only derive the association,but also indicate the degree of association.