Research On The Visual Association Rules Mining Based On Patching Item

As the swift development of computer hardware and information technology,how to process massive data has become an important work in researching and manufacture fields,so the technology of data mining came into being.At the same time,how to help users and analysts to discover interesting information more quickly and more intuitively or to provide them with more novel solution became to be another question.Fortunately, visualization technology always is a useful tool to this kind of questions.So,research of applying visualization technology into data mining is a job of meaningful.Applying visualization technology to data mining is helpful for users to understand the mining process,get the mining result and make a decision more quickly.Different mining pattern require different algorithms and different visualization technologies arc also needed.Visual data mining is a hotspot question in the data mining field.Based on the research results at home and abroad,we choose the following as two of our research directs:Our research focus on algorithms of association rules,especially the algorithms of mining multi-level association rules.We also notice the problem of redundancy existing in multi-level structure,and propose a redundancy deleting method and applying it to the algorithm based on patching item.We prove that the algorithm is improved bi experiment..We also compare several visualization methods and discuss their advantages and disadvantages.We design a visualization method to indicate association rules intuitively, in which,we have already thought of the interactively between users and the computer. In the last,we describe each step of a visualization system and realize them.