An object-oriented approach to multi-level association rule mining

Abstract

Previous studies in data mining have yielded efficient algorithms for discovering association rules. To increase the expressiveness and relevance of the discovered rules, researche~s typically augment the thresholds with domain-specific knowledge. Most existing proposals use the relational approach to organize and maintain these multi-level concept hierarchies. We argue that an object-oriented approach is better suited for focusing the search and regulating the mining of association rules both at multi-levels within one concept hierarchy, and across multiple concept hierarchies. We propose an adaptive encodhtg scheme for focusing the mining on semantically deeper and more informative knowledge.We demonstrate that the application of an object-oriented approach provides us with the benefits of a flexible combination of multiple multi-level concept hierarchies for focusing the mining on more informative and refined knowledge, and also allows us to enjoy seamless integration of multi-level concept hierarchies with legacy databases. In addition, by using an adaptive encoding scheme, efficient algorithms developed for discovering association rules can be integrated into the object-oriented framework with no or little extm cost.

A Framework for Object-Oriented Data Mining Based on Higher-Order Logic Programming

“…Fortin and Liu presented an object-oriented approach to multi-level association rule mining [5,6]. Association rules are mined in the context of multi-level concepts hierarchies which are represented by hierarchy of classes.…”

Using an Interest Ontology for Improved Support in Rule Mining

“…Fortin et al [3] use an object-oriented representation for data mining. Their interest is in deriving multi-level association rules.…”