RELigator: Chemical-disease relation extraction using prior knowledge and textual information

Pons, Ewoud; Becker, Benedikt FH; Akhondi, Saber; Afzal, Zubair; van Mulligen, Erik M.; Kors, Jan
Abstract:
The Erasmus MC team participated in the chemical-disease relation (CDR) task in the BioCreative V challenge. The CDR task consists of two subtask: automatic disease named entity recognition and normalization (DNER) and extraction of chemical-induced diseases (CID) from Medline abstracts. For the DNER subtask, we used our concept recognition tool Peregrine, in combination with several optimization steps. For the CID subtask, our system – RELigator – was trained on a rich feature set, including features derived from a graph database containing prior knowledge about chemicals and diseases, and linguistic and statistical features derived from the training corpus abstracts. We describe the systems that we developed and used, provide evaluation results for both CDR subtasks on the reference set, and compare the performance of our systems with baseline systems provided by the challenge organizers.
Year:
2015
Type of Publication:
Article
Journal:
Proceedings of the Fifth BioCreative Challenge Evaluation Workshop
Volume:
1
Pages:
247-253
Hits: 890