Title: An Inquiry into Question-Answering and Deep QA
The objective of the presentation is to explore question-answering as an autonomous research field. Beyond the basic definitions, we attempt to revisit some of the fundamentals in the field. In particular, we explore the role played by dimensions such as domain specificity, corpus specificity and size, natural language understanding, and macro-/micro-reading. We survey the (recent) history of question-answering and introduce some of the (rare) existing systems. To support our demonstration, we capitalize on more than a decade of effort to develop the EAGLi question-answering framework, a testbed used to explore several question-answering tasks, including definitional, factoid and deep Question-Answering.
Patrick Ruch is currently a professor in the Information Sciences department of the University of Applied Sciences Western Switzerland (HES-SO) in Geneva. He is also leader of the Text Mining group of the SIB Swiss Institute of Bioinformatics, which maintains text analytics services to support the curation of SIB databases such as Swiss-Prot and neXtprot (~ one million queries a year). Before these appointments, he has been working for a dozen years in several clinical & research environments, including the National Library of Medicine (Bethesda, MD), the University Hospitals of Geneva, and IBM Research in Zürich, where he developed text-driven decision support systems for life scientists and healthcare professionals.