Inferring knowledge from frequent patterns

Abstract

Many knowledge discovery problems can be solved efficiently byfmeans of frequent patterns present in the database. Frequent patterns are usefulfin the discovery of association rules, episode rules, sequential patterns andfclusters. Nevertheless, there are cases when a user is not allowed to access thefdatabase and can deal only with a provided fraction of knowledge. Still, the userfhopes to find new interesting relationships. In the paper, we offer a new methodfof inferring new knowledge from the provided fraction of patterns. Two newfoperators of shrinking and extending patterns are introduced. Surprisingly, afsmall number of patterns can be considerably extended into the knowledgefbase. Pieces of the new knowledge can be either exact or approximate. In thefpaper, we introduce a concise lossless representation of the given and derivablefpatterns. The introduced representation is exact regardless the character of thefderivable patterns it represents. We show that the discovery process can befcarried out mainly as an iterative transformation of the patterns representation.