Towards Discovery of Information Granules

The amount of electronic data available is growing very fast and this explosive growth in databases has generated a need for new techniques and tools that can intelligently and automatically extract implicit, previously unknown, hidden and potentially useful information and knowledge from these data. Such tools and techniques are the subject of the field of Knowledge Discovery in Databases. Information granulation is a very natural concept, and appears (under different names) in many methods related to e.g. data compression, divide and conquer, interval computations, neighborhood systems, and rough sets among others. In this paper we discuss information granulation in knowledge discovery. The notions related to information granules are their syntax and semantics as well as the inclusion and closeness (similarity) of granules. We discuss some problems of KDD assuming knowledge is represented in the form of information granules. In particular we use information granules to deal with the problem of stable (robust) patterns extraction.