Analysis And Research On Internet User Behavior Based On Interest Mining

Since the 21st century,with the rapid development of computer technology,the era of mobile Internet has quietly arrived.As a branch of the field of artificial intelligence,data mining has received extensive attention and research,and has formed different research directions such as cluster analysis,association rules,and classification.Association rule mining is an important part of the data mining field,and its purpose is to discover valuable patterns between transactions from massive data.However,the association rule mining algorithm using the support-confidence mining framework can easily produce misleading rules and ignore some of the rules with low support and high relevance.In order to solve this problem,an interest measurement model is introduced to verify whether the mining results are really valuable to users.Aiming at the shortcomings of the traditional association rule mining framework,this paper analyzes the commonly used interest measurement models.For the problems of item set symmetry and measurement value in the objective interest degree,the subjective interest degree measurement is introduced,and an interest degree fusion model based on the combination of subjective and objective interest degree is proposed.In addition,through the study of positive and negative association rules algorithms,it is found that although researchers have proposed mining algorithms for positive and negative association rules from different perspectives,there is still a problem of rule explosion when mining negative association rules.In this paper,by introducing the definition of Max-support to limit the generation of itemsets when mining frequent itemsets,this paper proposes an algorithm for mining positive and negative association rules based on the measure of interest.The algorithm is verified by two sets of real data sets.Experiments show that the IntRIMine algorithm can effectively mine positive and negative association rules and reduce the generation of useless rules and misleading rules.Then,this paper uses the proposed interest-based positive and negative association rule mining algorithm to mine network user behavior data,and find valuable rules in it.After data preprocessing,DC-clique,an improved clique clustering algorithm is proposed through the research of high-dimensional subspace clustering algorithm.Use this algorithm to cluster user data,divide users with the same characteristics into the same cluster,and mine association rules in the cluster to improve the efficiency of mining.Then according to the processed data,mining association rules.According to the experimental results,the analysis of network user behaviors verifies the effectiveness and applicability of the proposed interest model and association rule mining algorithm on the one hand,and on the other hand,it also provides certain guidance for website users and platform operations.