Publication detail

Multi-level Sequence Mining Based on GSP

ŠEBEK, M. HLOSTA, M. KUPČÍK, J. ZENDULKA, J. HRUŠKA, T.

Original Title

Multi-level Sequence Mining Based on GSP

English Title

Multi-level Sequence Mining Based on GSP

Type

conference paper

Language

en

Original Abstract

Mining sequential patterns is an important problem in the field of data mining and many algorithms and optimization techniques have been published to deal with that problem. An GSP algorithm, which is one of them, can be used for mining sequential patterns with some additional constraints, like gaps between items. Taxonomies can exist upon the items in sequences. It can be applied to mine sequential patterns with items on several hierarchical levels of the taxonomy. If a more general item appears in a pattern, the pattern has higher or at least the same support as the one containing the corresponding specific item. This allows us to mine more patterns with the same minimal support parameter and to reveal new potentially useful patterns. This paper presents a method for mining multi-level sequential patterns. The method is based on the GSP algorithm and generalization of more specific sequences based on the information theory.

English abstract

Mining sequential patterns is an important problem in the field of data mining and many algorithms and optimization techniques have been published to deal with that problem. An GSP algorithm, which is one of them, can be used for mining sequential patterns with some additional constraints, like gaps between items. Taxonomies can exist upon the items in sequences. It can be applied to mine sequential patterns with items on several hierarchical levels of the taxonomy. If a more general item appears in a pattern, the pattern has higher or at least the same support as the one containing the corresponding specific item. This allows us to mine more patterns with the same minimal support parameter and to reveal new potentially useful patterns. This paper presents a method for mining multi-level sequential patterns. The method is based on the GSP algorithm and generalization of more specific sequences based on the information theory.

Keywords

Sequence pattern mining, GSP, taxonomy

RIV year

2011

Released

16.11.2011

Publisher

Faculty of Electrical Engineering and Informatics, University of Technology Košice

Location

Košice

ISBN

978-80-89284-94-8

Book

Proceedings of the Eleventh International Conference on Informatics INFORMATICS'2011

Edition

1

Edition number

NEUVEDEN

Pages from

185

Pages to

190

Pages count

6

Documents

BibTex


@inproceedings{BUT76372,
  author="Michal {Šebek} and Martin {Hlosta} and Jan {Kupčík} and Jaroslav {Zendulka} and Tomáš {Hruška}",
  title="Multi-level Sequence Mining Based on GSP",
  annote="Mining sequential patterns is an important problem in the field of data mining
and many algorithms and optimization techniques have been published to deal with
that problem. An GSP algorithm, which is one of them, can be used for mining
sequential patterns with some additional constraints, like gaps between items.

Taxonomies can exist upon the items in sequences. It can be applied to mine
sequential patterns with items on several hierarchical levels of the taxonomy. If
a more general item appears in a pattern, the pattern has higher or at least the
same support as the one containing the corresponding specific item. This allows
us to mine more patterns with the same minimal support parameter and to reveal
new potentially useful patterns. This paper presents a method for mining
multi-level sequential patterns. The method is based on the GSP algorithm and
generalization of more specific sequences based on the information theory.",
  address="Faculty of Electrical Engineering and Informatics, University of Technology Košice",
  booktitle="Proceedings of the Eleventh International Conference on Informatics INFORMATICS'2011",
  chapter="76372",
  edition="1",
  howpublished="print",
  institution="Faculty of Electrical Engineering and Informatics, University of Technology Košice",
  year="2011",
  month="november",
  pages="185--190",
  publisher="Faculty of Electrical Engineering and Informatics, University of Technology Košice",
  type="conference paper"
}