Afternoon, all!
Myself and the other IS coordinators are looking at setting a deadline for the 2020A edition TREC-IS in early to mid March. The current thought is to release a larger dataset of tweets around 15 events, have participants run their systems against these events in a manner similar to the prior track editions, collect results, and then we’ll have annotators evaluate a subset of tweets submitted using pooling.
The big question at the moment is whether that timeframe of early- to mid-March is too soon for participants. Anyone have a concern with that timeframe?