Factored Sequence Kernels
Pierre Mahé, Nicola Cancedda
In this paper we propose an extension of sequence kernels to the case where the symbols that define the sequences have multiple representations. This configuration occurs, for instance, in natural language processing, where words can be characterized according to different linguistic dimensions. The core of our contribution is to integrate early the different representations in the kernel, in a way that generates rich composite features defined across the various symbol dimensions.
Appearing on the Journal Neurocomputing, Volume 72, Issues 7-9, March 2009, Pages 1407-1413. The article is available online