CLEF NewsREEL 2016

Author: Benjamin Kille

We are happy to announce that NewsREEL has been accepted for CLEF 2016. NewsREEL is short for “News Recommendation Evaluation Lab”. NewsREEL offers two tasks for researchers to engage with.

First, we provide a data set comprising interactions between users and news articles on a variety of digital publisher services. Participants re-iterate these event streams and evaluate strategies to suggest news articles. In the scope of CrowdRec – a collaborative project funded by the EU – we developed Idomaar. Idomaar is a reference framework for recommender system evaluation. Idomaar allows participants to evaluate streams. Thereby, participants can focus on their recommendation algorithm. Idomaar takes care of the technicalities. In addition, Idomaar reports technical quality measures, such as throughput, execution time and errors. Hence, we obtain a better picture of algorithms’ space and time complexity.

Second, we offer direct access to a real-time news recommendation service. The service is operated by plista and accessible via ORP (open recommendation platform). Unlike data sets, ORP allows participants to create value for actual users. We observe algorithms’ performance directly. As we continue to observe click-through-rates, we obtain meaningful comparisons between algorithms.

We are looking forward to your participation. Registration opens on 2 November 2015. We would like to keep you posted. You can either visit the NewsREEL website or follow us on Twitter (@CLEFNEWSREEL). If you happen to be at RecSys this week, feel free to talk to me or grab one of our flyers:

NewsREEL 2016 Flyer

Leave a Reply

You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>