Research And Application Of An Improved Algorithm For Association Rules

With the rapid and thorough development of data mining technology,more and more scholars and researchers pay attention to the association rules and related technologies.Mining association rules aims at finding interesting correlations and associations from big volumes of data.The background of the application expands from the original shopping basket analysis to the network intrusion detection,the user consumption habits analysis,the association rule classification,the traffic accident pattern analysis,the software bug mining and so on.Therefore,it is of great practical significance to study the technology of association rules.In this paper,we introduce the basic concepts,research status.,application and development trend of data mining at first.Then analyses and summarize some classical algorithms of association rule On this basis,a new association rule mining algorithm based on maximal frequent itemsets is proposed:Firstly,the MFIP-Miner algorithm for mining maximal frequent itemsets is proposed.The algorithm makes full use of the frequent pattern tree(FP-Tree)to compress the transaction in the database,and uses the feature of the frequent pattern tree.It does not need to generate candidate itemsets in the process of mining maximal frequent itemsets.Secondly,based on the theoretical analysis of MFIP-Miner algorithm,this paper completes the experimental platform,using the language of R,in the Eclipse + StatET programming environment to achieve the algorithm,and completes the comparison between the algorithms.Finally,the MFIP-Miner algorithm is applied to the weather forecast system,and the design and implementation of the forecasting system is completed.