About UCASTER

InstallProfileTry ItAbout

Video

UCASTER in action

What happens when you click the UCASTER icon?

interesting details about UCASTER

  1. The text on your page is sent to a service that analyzes it and detects the key topics by combining results from multiple semantic analysis and natural language processing algorithms.
  2. UCASTER accesses the Twitter search API, the NSDL Data Repository search API, and a small repository of "apps" (selected for the demo).
  3. UCASTER searches NSDL using narrow searches on the key topics restricted to the given grade levels. If not enough results are returned, the searches are broadened. In the demo there is no re-ranking, optimization, or link checking.
  4. The tweet stream from the designated account and its followers is fed into an algorithm that uses latent semantics to determine relevance. The most relevant tweets are displayed.
  5. We have created and meta-tagged a small set of weather apps, mostly selected from NSDL collections. UCASTER displays the best matches. 
  6. A "demo app” is always displayed first. It retrieves topical and profile information via a service. No other apps have been “contextualized.”

For details on the architecture and additional functionality please see the full NSF proposal.

What is NOT in this demo