String Kernels Based on Variable-Length-Don't-Care Patterns

Authors:
Kazuyuki Narisawa;Hideo Bannai;Kohei Hatano;Shunsuke Inenaga;Masayuki Takeda
Affiliations:
Department of Informatics, Kyushu University,;Department of Informatics, Kyushu University,;Department of Informatics, Kyushu University,;Graduate School of Information Science and Electrical Engineering, Kyushu University, Fukuoka, Japan 819---0395;Department of Informatics, Kyushu University,
Venue:
DS '08 Proceedings of the 11th International Conference on Discovery Science
Year:
2008

Citing 9
Cited 0

A training algorithm for optimal margin classifiers

COLT '92 Proceedings of the fifth annual workshop on Computational learning theory
Space-Economical Construction of Index Structures for All Suffixes of a String

MFCS '02 Proceedings of the 27th International Symposium on Mathematical Foundations of Computer Science
Discovering Best Variable-Length-Don't-Care Patterns

DS '02 Proceedings of the 5th International Conference on Discovery Science
The Minimum DAWG for All Suffixes of a String and Its Applications

CPM '02 Proceedings of the 13th Annual Symposium on Combinatorial Pattern Matching
Text classification using string kernels

The Journal of Machine Learning Research
Rational Kernels: Theory and Algorithms

The Journal of Machine Learning Research
Learning languages with rational kernels

COLT'07 Proceedings of the 20th annual conference on Learning theory
Unsupervised spam detection based on string alienness measures

DS'07 Proceedings of the 10th international conference on Discovery science
Finding patterns with variable length gaps or don’t cares

COCOON'06 Proceedings of the 12th annual international conference on Computing and Combinatorics

Quantified Score

Hi-index	0.00

Visualization

Abstract

We propose a new string kernel based on variable-length-don't-care patterns(VLDC patterns). A VLDC pattern is an element of (Σ茂戮驴 { 茂戮驴 })*, where Σis an alphabet and 茂戮驴 is the variable-length-don't-care symbol that matches any string in Σ*. The number of VLDC patterns matching a given string sof length nis O(22n). We present an O(n5 ) algorithm for computing the kernel value. We also propose variations of the kernel which modify the relative weights of each pattern. We evaluate our kernels using a support vector machine to classify spam data.