Gene Ontology concept recognition

In the past years the increasing usage of electronic publication has led to an enormous growth of databases and repositories containing articles relevant to scientifically different domains. Of these databases, the electronically available database PubMed [Putnam, 1998] has emerged as the most relevant one for the biomedical community. As of today, PubMed consists of more than 20 million abstracts [NLM Systems, 2011] and about 2 million full-text articles (accessible via PubMed Central [Vastag, 2000]) containing a huge amount of information on various biomedical topics. While PubMed is freely available, its size and rapid growth and the unstructured nature of text written in natural language call for tools and methods that assist in the process of information extraction in order to become more accessible and usable. Ontologies provide controlled vocabularies which may assist this process if a text passage can be mapped to an ontology concept.