Research Of Association Rules Mining Algorithm Based On Graph

As the rapid development of the computer technology,data mining technology is one of the most advanced research directions in Database and Data Warehouse fields.It is an effective approach to resolve the problem of abundant data and scanty information and it is currently the frontier in the information science field.The association rules mining is an important research subject in data mining field.It is widely used in each field which can both discover the hidden new rules and examine the knowledge pattern of long-term information inside the profession field.to discover effectively,comprehend and use the association rule is the important method to finish mining data.As new requirement continuously along with the market bring forward,data mining technology needs a high performance and dependable data mining algorithm.In the field of mining association rule.tiding maximal frequent itemsets is the key problem we must face.So our main goal is to find maximum frequent itemsets in transaction database,this paper mainly discuss how to mining maximum frequent itemsets effectively in database.This paper firstly introduces the basic conception,sorts and mission of Data Mining, and the defintion,mining procedures and kinds of association rules.Then,the paper expatiates on the classical association rules mining algorithms:Apriori algorithm and FP-growth algorithm.then analysize their application area,advantages and disadvantages, and make a comparison between them.The main parts of this paper refers to the research of association rules mining algorithm based on graph.to overcome the problems which exist in apriori algorithms, this paper presents improved IO-DLG algorithm based on existing association rules mining algorithms DLG,It was improved on three aspects which enhance the efficiency of mining frequent itemsets,then our purpose is achieved.