Welcome to the InfoSci Platform
Reference Hub15
Maximal Pattern Mining Using Fast CP-Tree for Knowledge Discovery

Maximal Pattern Mining Using Fast CP-Tree for Knowledge Discovery

R. Vishnu Priya, A.Vadivel, R. S. Thakur
Copyright: © 2012 |Volume: 3 |Issue: 1 |Pages: 19
ISSN: 1941-868X|EISSN: 1941-8698|EISBN13: 9781466612730|DOI: 10.4018/jissc.2012010106
Cite Article Cite Article

MLA

Priya, R. Vishnu, et al. "Maximal Pattern Mining Using Fast CP-Tree for Knowledge Discovery." IJISSC vol.3, no.1 2012: pp.56-74. http://doi.org/10.4018/jissc.2012010106

APA

Priya, R. V., A.Vadivel, & Thakur, R. S. (2012). Maximal Pattern Mining Using Fast CP-Tree for Knowledge Discovery. International Journal of Information Systems and Social Change (IJISSC), 3(1), 56-74. http://doi.org/10.4018/jissc.2012010106

Chicago

Priya, R. Vishnu, A.Vadivel, and R. S. Thakur. "Maximal Pattern Mining Using Fast CP-Tree for Knowledge Discovery," International Journal of Information Systems and Social Change (IJISSC) 3, no.1: 56-74. http://doi.org/10.4018/jissc.2012010106

Export Reference

Mendeley
Favorite Full-Issue Download

Abstract

The knowledge discovery from large database is useful for decision making in industry real-time problems. Given a large voluminous transaction database, the knowledge is discovered by extracting maximal pattern after some analysis. Various methods have been proposed for extracting maximal pattern including FP and CP trees. It has been noticed that time taken by these methods for mining is found to be large. This paper modifies tree construction strategy of CP-tree for mining maximal pattern and the strategy takes less time for mining. The proposed modified CP-tree is constructed in two phases. The first phase constructs the tree based on user given item order along with its corresponding item list. In the second phase, each node in the branch of the constructed tree is dynamically rearranged based on item sorted list. The maximal patterns are retrieved from the proposed tree using the FPmax algorithm. The proposed tree has been built to support both interactive and incremental mining. The performance is evaluated using both dense and sparse bench mark data sets such as CHESS, MUSHROOM, CONNECT-4, PUMSB, and RETAIL respectively. The performance of the modified CP-tree is encouraging compared to some of the recently proposed approaches.

Request Access

You do not own this content. Please login to recommend this title to your institution's librarian or purchase it from the IGI Global bookstore.