IHS-RD-BELARUS: Clinical Named Entities Identification in French Medical Texts

In this paper we present the results of our participation in the Task 1b of the 2015 CLEFeHealth challenge, whose goal was the identification of clinical entities of various types from medical texts in French and its normalization. We used the CRF-based system developed for disorder recognition in English and enhanced with French knowledge resources to recognize 10 types of clinic named entities from French medical texts: Anatomy, Chemical and Drugs, Devices, Disorders, Geographic Areas, Living Beings, Objects, Phenomena, Physiology and Procedures. Our system’s performance in entity recognition task was evaluated at 0.70 and 0.52 Fmeasure in exact match mode and 0.80 and 0.70 F-measure in inexact match mode depending on test corpus. The obtained results are higher than the average of all submitted runs.