Knowledge generalization from long sequence of execution scenarios
[ 1 ] Instytut Informatyki, Wydział Informatyki, Politechnika Poznańska | [ P ] pracownik
2013
referat
angielski
EN This paper describes study on generalization of information obtained from context sequential patterns. Problem definition of the context based sequential patterns mining extends the basic sequence structure and introduces heterogeneous attributes describing elements and sequences. Introduction of continuous context information requires knowledge aggregation after the mining. It is a result of many mined patterns that differ only on context attributes values. This paper introduces a new algorithm for the patterns generalization. It aggregates similar patterns and provides compact generalized patterns more readable for humans. Performed experiments shown that the algorithm provides reasonable generalization that accurately represents original knowledge mined from the database.
309 - 318
WoS (15)