home
past talks
resources
conferences
|
|
|
Relation extraction using tree kernels
Aron Culotta
UMass
Abstract
We extend previous work on tree kernels to estimate the similarity between the dependency trees of sentences. Using this kernel within a Support Vector Machine, we attempt to detect and classify relations between entities in the Automatic Content Extraction (ACE) corpus of news articles. The assumption is that entity relations share distinguishing phrasal structures that the tree kernel can capture. We examine the utility of different features such as Wordnet hypernyms, parts of speech, and entity types. We also explore ways of weighting such features to increase the kernel's effectiveness.
Back to ML Lunch home
|