Interactive Visualization of Search Intent for Exploratory Information Retrieval

Searching for relevant documents from a vast amount of scientific data is a challenging problem that requires a close interaction between the user, the search interface and the search engine. This extended abstract summarizes recent research on Intent Radar, an interactive search user interface that allows the user to directly interact with her estimated search intent, and in this way direct the search in an intuitive way without the need to type specific queries. In user experiments, Intent Radar improves task performance and quality of retrieved information without compromising the task execution time. This workshop paper presents to the ICML Crowdsourcing and Human Computing workshop audience our work recently published in CIKM 2013 and IUI 2013 as well as a short discussion of ongoing work on the topic. ICML ’14 Workshop: Crowdsourcing and Human Computing, Beijing, China, 2014. Copyright 2014 by the author(s).