Speaker:
Title:
Big Data in Search and Beyond: By People, For People
Abstract:
People generate an ever-growing amount of behavioral data when they interact with computer systems. Rather than treating these data purely as numbers or tokens, I will present projects that decode user behavior from the data and construct practical models. One project collects mouse cursor activity on a live search engine, and incorporates these data in two user models: one to understand visual attention to predict where people are looking without an eye-tracking device, and a graphical model that can be used to improve the relevance of search results. I will provide examples of how these methods can be applied to studying users interacting with games, mobile devices, reviews, and the web. Through this, we can better understand fundamental human behavior and help design systems that allow people to find information faster and easier.
Short Bio: Jeff Huang is a PhD Candidate in Information Science at the University of Washington. His research in data-driven information retrieval focuses on modeling users from interaction data. He has been awarded Best Paper at SIGIR 2010, Honorable Mention at CHI 2011, and the Facebook Fellowship. During his graduate studies, Jeff has conducted research at the University of Washington and five research groups at Microsoft Research and Google, and has received external funding from Google and Microryza. His work appears in venues such as SIGIR, CHI, AAAI, UIST, CIKM, WSDM, as well as in the Wall Street Journal, GeekWire, and the MIT Technology Review. Jeff earned his Masters and Bachelors degrees in Computer Science at the University of Illinois.