DiscoverText Analysis from the State of the Union

One of the partners in the SurveySwipe State of the Union mobile study experiment was DiscoverText – a company that does intelligent text analysis of extremely large datasets (i.e. “What are people saying about the State of the Union on Twitter?”)

DiscoverText has released the raw datasets from the text analytics portion of the study. Dr. Stuart Shulman is still working on the visualization of such a large dataset. We’ll provide updates here as his analysis becomes available. In the meantime, the raw data is available at http://discovertext.com/sotu.aspx.

DiscoverText aims to reinvent Text Analytics

I got introduced to Stu Shulman – a Political Science Professor out of UMASS-Amherst through a common friend in DC – as I was making the rounds in DC for our IdeaScale Federal Government business. I got hooked to what he was doing with Text Analytics and the sheer volume of research he has done within the academic community and how he’s going about bringing all that research (funded by your and my tax dollars) to the commercial space. Stu is also the CEO of DiscoverText and is taking the plunge into entrepreneurship. He is also the Editor-in-Chief of the Journal of Information Technology & Politics.

DiscoverText was just launched on November 1st, and I think it has the power to make a dent in the text analytics marketplace. I’ve been in the feedback space for quite some time, and I can tell you from first hand experience, open-ended text analytics and sentiment analysis is not an easy problem to solve. In fact many tools claim to do it, and very few are good at it. More importantly text analytics has to be effective. With the sheer volume of text data on the internet today (tweets, facebook updates, blog comments) – this is a treasure trove of data that is waiting to be mined.
[Read more...]