Title :
Pattern Mining as Abduction: From Snapshots to Spatio-Temporal Sequential Patterns
Author :
Hazarika, Shyamanta M.
Author_Institution :
Tezpur Univ., Tezpur
Abstract :
The focus of this paper is towards development of a logical framework for mining spatio-temporal sequential patterns. The spatial representation language RCC-8, often referred to as region connection calculus and its spatio-temporal extension, ST0, a fragment of prepositional spatio-temporal logic is used as the knowledge representation formalism. Standard abduction is used for mining sequential patterns in spatio-temporal data. Abductive reasoning may yield more than one possible answer and is accompanied by some preference criteria expressed using heuristics. Here, the abduction technique is circumscription which implements the heuristic that changes should only occur when forced to. Various additional heuristics to drive the selection of preferred explanations are discussed.
Keywords :
data mining; knowledge representation languages; pattern mining; spatial representation language; spatio-temporal sequential patterns; Application software; Calculus; Computer vision; Conferences; Data mining; Drives; Knowledge representation; Logic programming; Spatial databases; Transaction databases;
Conference_Titel :
Data Mining Workshops, 2007. ICDM Workshops 2007. Seventh IEEE International Conference on
Conference_Location :
Omaha, NE
Print_ISBN :
978-0-7695-3019-2
Electronic_ISBN :
978-0-7695-3033-8
DOI :
10.1109/ICDMW.2007.71