Studies And Applications Of Association Rule Mining Methods In Data Mining

"Data explosion and knowledge scarcity"is an urgent phenomenon in information society. However, data mining is the most effective method to tackle the problem, which is a process of extracting useful information and identifying valid, novel, potentially useful, and ultimately understandable patterns in data from large volumes of raw data in order to solve the practical problems. Therefore, study on data mining technique is of important practical meaning. The dissertation focuses on mining methods of association rule in data mining and gives a detailed analysis of association rule mining algorithms.The work of author mainly focuses on four aspects in the following: 1. Constrained maximal frequent itemsets mining algorithm (CMFS) is proposed in order to mine information with emphasis and purpose according to users'demands; 2. An incremental maintaining algorithm for frequent closed itemsets (UCHARM) is proposed in order to utilize mined useful information to fast realize maintenance of frequent closed itemsets; 3. In order to solve reasonable attribute partition domain in classical quantitative association rule mining algorithm, a qualitative quantitative conversion model, cloud model is deeply studied. The mechanism of uncertainty reasoning and various cloud generators algorithms are deeply studied; 4. An interactive visualization method is developed in order to help users more clearly and easily understand meaning of association rules. A lot of numerical examples are given to validate the new algorithms proposed in each chapter in the dissertation.Association rule mining algorithms developed in the dissertation are applied in database forecasting project, i.e. analysis system of telephone sheet on Tetong platform. By mining association rules in telephone sheet data, potential and valuable information will be found.