Abstract

Biomedical natural language processing (BioNLP) has received a lot of 
attention in recent years. Due to the lack of annotated material, as 
well as the differences from the newswire domain on which most NLP 
technology has been developed, the biomedical domain represents a 
realistic test for the portability and adaptation of existing methods 
and approaches.

In the context of the FlySLIP project, we developed NLP components 
which are used to support the curation of the FlyBase database. In 
this talk, I am going to concentrate on our work on biomedical named 
entity recognition and disambiguation and how we overcame the lack of 
annotated material. Also, I will discuss annotation and evaluation issues.