Detección de fármacos genéricos en textos biomédicos
- Segura Bedmar, Isabel
- Martínez Fernández, Paloma
- Samy, Doaa
ISSN: 1135-5948
Year of publication: 2008
Issue: 40
Pages: 27-34
Type: Article
More publications in: Procesamiento del lenguaje natural
Abstract
This paper presents a system for drug name recognition and clasification in biomedical texts. The system combines information from UMLS Metathesaurus and nomenclatura rules for generic drugs, recommended by United States Adoptated Names (USAN), that allow the classification of the drugs in pharmacologic families. The initial hypothesis is that rules are able to detect possible candidates of drug names which are not included in the UMLS database (version 2007AC), increasing, in this way, the coverage of the system. The system achieves a 100% precision and 97% recall using UMLS only. The combination of the USAN rules and UMLS slightly improves the coverage of the system.