Institutional Repository
Technical University of Crete
EN  |  EL

Search

Browse

My Space

SPIRIT: Sequential pattern mining with regular expression constraints

Garofalakis Minos, Rastogi Rajeev, Shim Kyuseok

Full record


URI: http://purl.tuc.gr/dl/dias/B103115A-82D8-4F23-BF34-145A4765DDA5
Year 1999
Type of Item Conference Publication
License
Details
Bibliographic Citation M.N. Garofalakis, R. Rastogi and K. Shim, "SPIRIT: Sequential pattern mining with regular expression constraints", in 25th VLDB Conference, September 1999, pp. 223-234.
Appears in Collections

Summary

Discovering sequential patterns is an important problem indata mining with a host of application domains includingmedicine, telecommunications, and the World Wide Web.Conventional mining systems provide users with only avery restricted mechanism (based on minimum support)for specifying patterns of interest. In this paper, we proposethe use of Regular Expressions (REs) as a flexibleconstraint specification tool that enables user-controlledfocus to be incorporated into the pattern mining process.We develop a family of novel algorithms (termed SPIRIT– Sequential Pattern mIning with Regular expressIon consTraints)for mining frequent sequential patterns that alsosatisfy user-specified RE constraints. The main distinguishingfactor among the proposed schemes is the degreeto which the RE constraints are enforced to prune thesearch space of patterns during computation. Our solutionsprovide valuable insights into the tradeoffs that arisewhen constraints that do not subscribe to nice properties(like anti-monotonicity) are integrated into the mining process.A quantitative exploration of these tradeoffs is conductedthrough an extensive experimental study on syntheticand real-life data sets.

Services

Statistics