Abstract:

The primary mission of incremental updating association rule mining is to solve the problem of the maintenance of the association rules, when transaction records are updated constantly and minimum support is changed in the online database. Given that many incremental update association rule mining algorithms have many problems, such as low efficiency and high computation cost, as well as difficult maintenance of the association rules, we propose an efficient algorithm for incremental updating association mining based on inverted index tree. The algorithm effectively combines the inverted index and the tree structure, so that the data in the transaction database is updated constantly with time and the minimum support degree is changed continuously with the application environment. Frequent item sets are generated without scanning the original transaction database and without generating candidate item sets. Experimental results show that the proposed algorithm can effectively solve the problem of incremental updating association rules with less memory space and higher efficiency.

Key words: incremental updating mining;inverted index;inverted index tree;frequent item sets;association rules