Proceedings of the 24th Conference on Computational Natural Language Learning

The 2005 Conference on Computational Natural Language Learning (CoNLL-2005) is the ninth in a series of meetings organized by SIGNLL, the ACL special interest group on natural language learning. This year's CoNLL will be held in Ann Arbor, Michigan, on June 29 and 30, in conjunction with the ACL 2005 conference. This year we encouraged submissions addressing "deeper linguistic phenomena than have typically been covered in the past", a theme reflected in the shared task, and one which will be addressed by our invited speakers, Mark Johnson and Mark Steedman. The latter talk will be part of what should be a very interesting joint session with the Workshop on Psychocomputational Models of Human Language Acquisition organized by William Sakas. A total of 70 papers were submitted to CoNLL's main session, of which we were able to accept only 19, making this the most competitive CoNLL meeting to date. We are particularly grateful to our program committee for their work reviewing this large number of papers on a tight schedule. In keeping with the unique tradition of CoNLL, we also have a shared task session this year, on semantic role labeling. This year's task included use of full syntactic parses, a step beyond the chunk-level information used in 2004. The shared task was coordinated by Xavier Carreras and Lluis Marquez. Common training and test data were made available to all participants, who had to build their learning system and test it on this task. The shared task also achieved a record number of submissions this year, and these proceedings include system descriptions from each of the 19 participants.