Rapid prototyping of pattern mining problems isomorphic to boolean lattices

• Main Authors: Flouvat, F., De Marchi, F., Petit, J.-M.
• Format: Conference Proceeding
• Language: English
• Subjects: Algorithm design and analysis; Data mining; Data structures; Itemsets; Lattices; Libraries; pattern; toolkit
• ISSN: 2151-1349; 2151-1357
• DOI: 10.1109/RCIS.2008.4632104

Interesting pattern mining is an important family of data mining problems with applications in many domains. In this paper, we focus on the special class of pattern mining problems known to be dasiarepresentable as setspsila. The main contribution of this paper is to take advantage of the common theoretical background of these problems from an implementation point of view by providing efficient data structures for boolean lattice representation and several implementations of well known algorithms. By the way, these problems can be implemented with only minimal effort, i.e. programmers do not have to be aware of low level code, customized data structures and algorithms being available for free. A toolkit, called iZi, has been devised and applied to several problems such as itemset mining, constraint mining in relational databases and query rewriting in data integration systems. According to our first results, the programs obtained using our toolkit offer a very good tradeoff between performances and development simplicity. Some methodological guidelines are also provided to guide the programmers both at the theoretical level and at the code level.