Deep Learning in Biomedical Text Mining: Contributions and Challenges

A large number of biomedical texts are published every day in scientific literature. Finding the relevant and useful information from the massive collection of scientific literature is a challenging task that can be compared to finding needles in the haystack. Biomedical text mining is one of the sophisticated methodologies that leverage the extraction of knowledge from existing biomedical texts automatically. Deep learning (DL) based techniques have rejuvenated this field with huge prospects. In this chapter, we highlighted the contribution of DL based techniques in three specific tasks in the field of biomedical text mining: named-entity recognition, relationship extraction, and question answering. We also discussed the DL based models that are proven to be successful in multiple natural language processing tasks and the related challenges we face using such DL based techniques. We believe DL based methods will play a significant role in the coming years for biomedical text mining.