Coreference Resolution in Biomedical Texts: a Machine Learning Approach

Abstract

Motivation: Coreference resolution, the process of identifying different
mentions of an entity, is a very important component in a
text-mining system. Compared with the work in news articles, the
existing study of coreference resolution in biomedical texts is quite
preliminary by only focusing on specific types of anaphors like pronouns
or definite noun phrases, using heuristic methods, and running
on small data sets. Therefore, there is a need for an in-depth
exploration of this task in the biomedical domain.
Results: In this article, we presented a learning-based approach
to coreference resolution in the biomedical domain. We made three
contributions in our study. Firstly, we annotated a large scale coreference
corpus, MedCo, which consists of 1,999 medline abstracts
in the GENIA data set. Secondly, we proposed a detailed framework
for the coreference resolution task, in which we augmented the traditional
learning model by incorporating non-anaphors into training.
Lastly, we explored various sources of knowledge for coreference
resolution, particularly, those that can deal with the complexity of
biomedical texts. The evaluation on the MedCo corpus showed promising
results. Our coreference resolution system achieved a high
precision of 85.2% with a reasonable recall of 65.3%, obtaining an
F-measure of 73.9%. The results also suggested that our augmented
learning model significantly boosted precision (up to 24.0%) without
much loss in recall (less than 5%), and brought a gain of over 8% in
F-measure.