Biomedical named entity recognition And drug drug association Extraction from literature with Deep neural networks

Abstract

Bioinformatics is a field of computational science concerned with the acquisition, storage, analysis, and dissemination of biological data. Bioinformatics uses computer programs for a variety of applications, including extracting information from the text and analyzing the association between them. Biomedical Named Entity Recognition (BNER) is the process of identifying the biomedical terms in the text and this is essential for biomedical information extraction. In order to improve information access on chemical compounds and chemical entities mentioned in the research documents, it is very crucial to be able to identify chemical entity mentions automatically within the text. Also, the simultaneous administration of too many drugs on the same patient has become considerably increased. The patients should be aware of the medications, say adverse or guidance to protect them from other effects during the prescribing process. So, identifying Drug-Drug Interactions (DDI) automatically from the biomedical related text is much needed. Even though many algorithms are available in the literature on the areas of biomedical named entity recognition and DDI association extraction, the use of a large feature set, biomedical dictionary, and framing of rules make the process of identifying the biomedical entities from the literature text more challenging. Moreover, recognizing the biomedical entities and identifying the DDI association over the long-term dependencies from the literature affects the performance, since most of the existing systems failed to recognize the entities and their relations newline newline newline