Kernel Functions

Research on document similarity has shown that complex representations are not more accurate than the simple bag-of- words. Term clustering, e.g. using latent semantic indexing, word co-occurrences or synonym relations …...

We propose an approach for extracting re- lations between entities from biomedical literature based solely on shallow linguis- tic information. We use a combination of kernel functions to integrate two …...

Various types of structural information - e.g., about the type of constructions in which binding constraints apply, or about the structure of names - play a central role in coreference …...

Relation extraction is the task of recog- nizing semantic relations among entities. Given a particular sentence supervised ap- proaches to Relation Extraction employed feature or kernel functions which usu- ally …...

AbstractProminently used in support vector machines and logistic re-gressions, kernel functions (kernels) can implicitly map data points into high dimensional spaces and make it easier to learn complex decision boundaries. …...