UMCC_DLSI: Semantic and Lexical features for detection and classification Drugs in biomedical texts

Second Joint Conference on Lexical and Computational Semantics (*SEM), Volume 2: Seventh International Workshop on Semantic Evaluation (SemEval 2013), pages 636–643, Atlanta, Georgia, June 14-15, 2013. c⃝2013 Association for Computational Linguistics UMCC_DLSI: Semantic and Lexical features for detection and classification Drugs in biomedical texts Armando Collazo, Alberto Ceballo, Dennys D. Puig, Yoan Gutiérrez, José I. Abreu, Roger Pérez Antonio Fernández Orquín, Andrés Montoyo, Rafael Muñoz Franc Camara DI, University of Matanzas Autopista a Varadero km 3 ½ Matanzas, Cuba {armando.collazo, dennys.puig, yoan.gutierrez, jose.abreu, roger.perez}@umcc.cu, alberto.ceballo@infonet.umcc.cu DLSI, University of Alicante Carretera de San Vicente S/N Alicante, Spain antonybr@yahoo.com, {montoyo, rafael}@dlsi.ua.es Independent Consultant USA info@franccamara.com Abstract In this paper we describe UMCC_DLSI- (DDI) system which attempts to detect and classify drug entities in biomedical texts. We discuss the use of semantic class and words relevant domain, extracted with ISR- WN (Integration of Semantic Resources based on WordNet) resource to obtain our goal. Following this approach our system obtained an F-Measure of 27.5% in the DDIExtraction 2013 (SemEval 2013 task 9). 1. Introduction To understand biological processes, we must clarify how some substances interact with our body and one to each other. One of these important relations is the drug-drug interactions (DDIs). They occur when one drug interacts with another or when it affects the level, or activity of another drug. DDIs can change the way medications act in the body, they can cause powerful, dangerous and unexpected side effects, and also they can make the medications less effective. As suggested by (Segura-Bedmar et al., 2011), “...the detection of DDI is an important research area in patient safety since these interactions can become very dangerous and increase health care costs”. More recent studies (Percha and Altman, 2013) reports that “…Recent estimates indicate that DDIs cause nearly 74000 emergency room visits and 195000 hospitalizations each year in the USA”. But, on the other hand, there is an expansion in the volume of published biomedical research, and therefore the underlying biomedical knowledge base (Cohen and Hersh, 2005). Unfortunately, as often happens, this information is unstructured or in the best case scenario semi-structured. As we can see in (Tari et al., 2010), “Clinical support tools often provide comprehensive lists of DDIs, but they usually lack the supporting scientific evidences and different tools can return inconsistent results”. Although, as mentioned (Segura-Bedmar et al., 2011) “there are different databases supporting healthcare professionals in the detection of DDI, these databases are rarely complete, since their update periods can reach up to three years”. In addition to these and other difficulties, the great amount of drug interactions are frequently reported in journals of clinical pharmacology and technical reports, due to this fact, medical literature becomes most effective source for detection of DDI. Thereby, the management of DDI is a critical issue due to the overwhelming amount of information available on them (Segura-Bedmar et al., 2011). 636 1.1. Task Description With the aim of reducing the time the health care professionals invest on reviewing the literature, we present a feature-based system …

Content truncated to 500 words