Summary: Lehrstuhl für Informatik 9
Zeit: Donnerstag, 2. Juli 2009, 16:30 Uhr
Ort: AH I, Ahornstr. 55
Referent: Vivekanand Gopalkrishnan, Ph.D.
Nanyang Technological University Singapore
Titel: Enhancing Utilities of Frequent Patterns
Frequent pattern mining is one of the most well-studied topics in data
mining. The set of frequent patterns of a transactional database can be
treated as a summary that contains all useful information within the data-
base. Previously, frequent patterns have been successfully applied for
classification, clustering, association rule mining, and indexing. This talk
covers two basic applications of frequent patterns: in database charac-
terization and in finding interesting patterns. We show the limitations of
these applications and discuss possible approaches to enhance them.
Because of the redundancy among the set of frequent patterns, it is usu-
ally too large for characterizing and understanding the database. To re-
solve this issue, we propose to summarize the frequent patterns in ways
that allow easier analysis.