Identifying and annotating generic drug names

This paper presents a system for identify generic drug names in biomedical texts. The identification of drug names is the first step of the process involved in the development of a system aimed at the automatic extraction of drug-drug interactions in biomedical texts. A drug-drug interaction occurs when one drug influences the level or activity of another drug. The proposed system uses the affixes, approved by the United States Adopted Name (USAN) Council, allow a specific classification of drugs on pharmacological families, and a linguistic engine based on finite-state techniques. As a preliminary approach, it is a first step towards a useful Information Extraction System in the field of Pharmacology. The results show that drug names are correctly identified. Our proposal will be extended to address additional information about pharmacologic classes.