Data Mining Frequent Set Algorithm

Data Mining, which is also called Knowledge Discovery in Database, has been widely researched these years in the database area. And Association Rules is one of the most important issues in Data Mining.Association rules is to fine out the relations among different items in the business database. These rules find out the behavior models that the customers have. For exemple, buying certain kind of goods has some influence towards buying another goods.These kinds of rules can be used in the area of designning storage rank, arranging the goods in stock, and catalog the customers by their behavior models.We first introduce some base knowledge in data mining, frequently-used methods of the data reduction, and the common algorithm of association rules. Then we discuss the relative theories in data warehouse in details. We discuss not only the objects and methods of data warehouse macroscopically, but also the implementations of the frequent itemset. And we also demonstrate the data structure and details, show the difference compared with other common algorithms, and point out the advantages.The FP-growth algorithm implemented in this article is a frequent itemset generation algorithm. Compared with other common frequent itemset generation algorithms, like Apriori algorithm, the advantages of FP-growth algorithm are that it doesn't need a large amount of candidate set, and it has better efficiency in both time complexity and space complexity. In the process of the implementation, we use the STL, Standard Template Library, of C++, and some data structures, such like vector, set and quick sort algorithm.This algorithm has some certain usages for data mining. On one hand, it offers some insights to the researchers of association rules, on the other hand, we can increase the query efficiency greatly, only if it were merged into other data mining tools.