Abstract
This paper proposes a novel approach to combining multiple rule sets for intelligent text mining. We develop a boosting-like technique for generating multiple sets of rules based on rough set theory and model decisions inferred from multiple rule sets as evidence for combination by Dempster's rule. This approach builds on our previous work on rough set based methods for mining maximal associations from text collections. Through evaluation on a benchmark data collection, the approach is demonstrated to provide significant improvement over single rule set methods and provides insight into incorporating evidential theory into text mining tasks.
Original language | English |
---|---|
Pages (from-to) | 191-209 |
Journal | Artificial Intelligence Review |
Volume | 26 |
Issue number | 3 |
DOIs | |
Publication status | Published (in print/issue) - Oct 2006 |
Bibliographical note
Other Details------------------------------------
Journal article awaiting print. This paper proposes a novel approach to combining multiple rule sets for intelligent text mining. We develop a boosting-like technique for generating multiple sets of rules based on rough set theory and model decisions inferred from multiple rule sets as evidence for combination by Dempster's rule. This approach builds on our previous work on rough set based methods for mining maximal associations from text collections. Through evaluation on a benchmark data collection, the approach is demonstrated to provide significant improvement over single rule set methods and provides insight into incorporating evidential theory into text mining tasks.
Keywords
- rough decisions
- text mining
- Dempster's rule