Question Answering

Question answering refers to the process of providing direct and precise answers to natural language questions. Biomedical question answering is a task directed towards aiding researchers, healthcare professionals and the public in managing the continuous growth of information in the biomedical domain. Question answering requires the use of complex natural language processing techniques in order to produce accurate responses. Question answering implies that the information about the topics of interest and the users’ preferences are extracted or inferred from the free-form questions posed in natural language. Questions and the need for personalized answers or summaries arise in all biomedical sub-domains: clinicians have questions about their patients that can be answered by the documents in electronic health records and the biomedical literature; patients have questions about their health that can be answered by consumer-friendly sources; biologists need informative summaries of the recent publications in their areas of interest; administrators have questions about healthcare policies; quality of healthcare, and disease outbreaks. These and many other areas of question answering and summarization are a fertile ground for research and development of applications. Note that many of the traditional knowledge-based approaches might seem irrelevant at the time of this writing as the Neural QA approaches show promise and potential to replace the traditional approaches. Knowing what was successful in the past and which elements are essential to getting the right answers, however, is needed to inform the development of the neural approaches. This chapter, therefore, provides an overview of the approaches to biomedical question answering as they were evolving.