Open Collections

Abstract

This thesis deals with a problem of mining sequential patterns from temporal databases. Data Mining is the non-trivial extraction of implicit, previously unknown, and potentially useful information from data. A sequential pattern, is an ordered list (ordered by increasing event's time) of events. The main contribution of this thesis is the development of a general model of sequential patterns as well as algorithms for discovering such patterns. We investigate the generalized sequential pattern in which time isn't just part of the constraints but also a part of the discovered pattern. We have developed several algorithms to discover the generalized sequential patterns from a given temporal database. We have implemented four of these algorithms in C++. The results we obtained from applying these implementations to various datasets are matching our expectations.