Abstract: Conversational information-seeking experiences (e.g., search, question answering) are a core functionality of modern virtual assistants. When a user is engaging with a particular topic (e.g., “What is the latest news about Travis Kelce and Taylor Swift?”), beyond the initial summary, they may wish to follow-up with additional information-seeking turns to better understand the associated topical context. In this talk, I will discuss our recent work regarding summarizing reported speech (e.g., “What did Patrick Mahomes say about their relationship?”), summarizing relevant background information (e.g., “Tell me about the history of their relationship.”), and incorporating information extraction techniques to both improve summaries and provide related entities for proactive exploration (e.g., “Would you like to know about the Chiefs season?”).
Bio: Kevin Small is a principal applied scientist in Amazon AI, working on developing conversational informational-seeking experiences (e.g., question answering, topic exploration). Before joining Alexa, Kevin has been a member of several teams within Amazon including Core Machine Learning, AWS, and Sponsored Products Advertising. His primary interests are natural language processing and machine learning with a particular interest in decision support and decision making systems. Before joining Amazon, Kevin received his PhD from the University of Illinois under the direction of Dan Roth and held academic positions at Tufts Medical Center and the National Institutes of Health.
