Effective Item Set Mining Using Incremental Update Strategy for IMine Index Structure

Abstract

Mining Item set with support constraints are widely used to discovery hidden knowledge from relational DBMS. To reduce the computational complexity of item set extraction, support constraint is enforced on the extracted item sets. Recent existing work, IMine index (Item set-Mine index), provided a structural data representation of transactional data to efficiently extract item set from a relational databases. However the growing demand for frequent updates in the database, necessitates the index structure to adapt to the modified transaction for extracting the item set effectively.

This paper presented an effective incremental update strategy to handle dynamic transaction of the databases for effective item set extraction. The incremental update strategy extract data item without accessing the original transactional, as the index created is not constrained with any support threshold. The update strategy handles the fast and frequent changing transaction and effective extract the data item compared to the existing Imine index structure. The index performance in terms of incremental updates is experimentally evaluated with data sets characterized by different size and data distribution. The experimental results shows better scalability of incremental update strategy for more frequently database updates in terms of transaction size and pattern length.