Semi-supervised learning to support the exploration of association rules

In the last years, many approaches for post-processing association rules have been proposed. The automatics are simple to use, but they don't consider users' subjectivity. Unlike, the approaches that consider subjectivity need an explicit description of the users' knowledge and/or interests, requiring a considerable time from the user. Looking at the problem from another perspective, post-processing can be seen as a classification task, in which the user labels some rules as interesting [I] or not interesting [NI], for example, in order to propagate these labels to the other unlabeled rules. This work presents a framework for post-processing association rules that uses semi-supervised learning in which: (a) the user is constantly directed to the [I] patterns of the domain, minimizing his exploration effort by reducing the exploration space, since his knowledge and/or interests are iteratively propagated; (b) the users' subjectivity is considered without using any formalism, making the task simpler. © 2014 Springer International Publishing.

Palavras-chave

Association Rules, Post-processing, Semi-supervised Learning (SSL)

Idioma

Inglês

Citação

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), v. 8646 LNCS, p. 452-464.