WARA Media and Language invites the WASP community to propose leaderboard challenges for machine learning. These challenges are essentially benchmarking tasks, where users submit solution attempts for automatic evaluation with respect to a set of criteria. The outcome is reported as a ranked list, i.e., a leaderboard, which is continiously updated as new solutions come in. Leaderboard challenges can, for example, be used to draw attention to a particular task or evaluation method, or as a teaching tool in ML courses.
To host the challenge with WARA ML arena, please provide (draft versions of) (i) a problem description, (ii) a target dataset, and (iii) an evaluation metric. Based on these, the WARA ML committee will decide on how to host the challenge. Proposals are submitted by email to nitheshk@cs.umu.se. More information is found in the linked pdf file below.