Odin: contextual document opinions on the go.

Information overload is a systemic problem for knowledge workers in enterprise. For a long time, information was scarce and therefore valuable. While, the explosion of digital information has made information plentiful, time to read and process that content is now scarce. This problem is only exacerbated by our increased mobility, and the expectation to be "on top" of the continuous barrage of documents while on the go. Knowledge workers in enterprise need solutions that are designed with quick methods for finding what to read in a large collection of documents (e.g. financial reports, legal documents, news), and ways of presenting it within small visual real estate. Unlike reviews, document collections are long, more varied, and context is extremely important. In response, we present Odin, a mobile web-based window onto a user's document corpus. Rather than performing corpus summarization, Odin users can quickly find opinions and documents that are Aligned or Divergent from the corpus' consensus, or those that are the most Relevant given the overall corpus' of opinions. Odin presents this information through a simple and intuitive mobile interface. To the authors' knowledge, this is the first UI/system (and support algorithm) to allow mobile users to place documents and their opinions in context through alignment rather than raw word count or sentiment. Positive results from two evaluations are also presented.