BayesFF in 45 seconds

I’m doing a 45 second presentation on the Bayes Feed Filter project at the JISC Rapid Innvation Development meeting in Manchester today. This is it:

The Bayesian Feed Filter will help researchers keep up to date with current developments in thier field. It will automatically filter RSS and ATOM feeds from Journals’ tables of content to (hopefully) select those that are relevent to an individual’s research interests.

It uses Bayesian statistical analysis, the same approach used in many spam filters. First you need to train it with samples of what you are and aren’t interested in; then it compares the frequency with which words occur in the text to predict whether new items are on a similar topic to the samples that you were interested in.

We are testing whether this approach works for researchers and Table of Content feeds and building an API, so would like to talk anyone who can use it to personalize their own data presentation.



Filed under dissemination

3 responses to “BayesFF in 45 seconds

  1. I’d be interested in keeping up with your progress on this project. I subscribe to a large number of RSS feeds. Finding the most interesting and relevant posts amongst the noise is time-consuming.

    Will the software be a stand-alone reader, a web service, a plugin for something like google reader or netvibes, or some other kind of implementation?

  2. Pingback: BayesFF: Final post « Bayesian Feed Filter