Technique Of Fuzzy Association Rule Mining And Its Application In The Recommended System

In the database of a large shopping site, there are various associations between the attributes of data,which is implict in the database.The aim is to find these hidden associations. The aim is to find these hidden associations. Meaningful association rules discovered from a large number of transaction records can help website managers to make business decisions, such as classification design, cross-shopping, thereby increasing sales and profits.In recent years, data mining and artificial intelligence technology becomes the hottest research area, which has attracted wide attention of experts and scholars. Fuzzy association rule mining is an important data mining technology and can be applied in a number of application areas.Meanwhile, the development of the Internet is facing a new challenge:on one hand, and the users are not completely interested in the mass product information provided by the site, and often need to go through multiple browser to find the products they want; on the other hand, the site does not fully understand the customer’s demand, and the users see the same interface. The product is not attractive, which lead to serious loss of customers. The most effective way is to personalize the user’s product recommendation.The main contents of this thesis are as follows:Firstly, based on the research results of association rule mining technology at home and abroad, it introduces the existing fuzzy association rule mining technology. Secondly, referring to the deficiencies of the algorithm for mining, it discusses a new fuzzy association rules mining algorithms, and compares the advantages of the algorithm by experimental analysis. Finally, taking the database shopping site for example, it makes research of fuzzy association rule mining technology which make full use of historical data in the database, and it also points out an effective form of mining fuzzy association rules, and studies new fuzzy association rules algorithm on personalized Recommender system.