Mining Segment-Wise Periodic Patterns in Time-Related Databases

Periodicity search, that is, searchfor cyclicity in time-related databases, is an interesting data mining problem. Mostprevious studies have been on finding full-cycle periodicity for all the segments in the selected sequencesof the data, that is, if a sequenceis periodic, all the points or segments in the period repeat. However,it is often usefill to minesegment-wiseor point-wiseperiodicity in time-related data sets. In this study, weintegrate data cube and Apriori data mining techniques for miningsegment-wiseperiodicity in regard to a fixed length period and showthat data cubeprovides an efficient structure and a convenient wayfor interactive miningof multiple-level periodicity.

• Type: Conference
• Year: 1998
• Where: KDD
• Authors: Jiawei Han, Wan Gong, Yiwen Yin
• Keywords: Data Mining, Data Mining Problem, Full-cycle Periodicity, KDD 1998