Abstract
Automatic access to drug target in biomedical articles is crucial for drug innovation. Determining a drug target requires specific context, which is usually expressed as a certain sentence. Thus, this study develops a two-stage cascading framework for extracting drug target from biomedical article, in which drug target sentences are identified first and then passed to the second stage to identify the drug target entity. The experiments show that our method could be a feasible way to automatically extract drug target knowledge from biomedical articles.