Proceedings of the Workshop on Multiword Expressions: Identifying and Exploiting Underlying Properties

This volume contains the papers accepted for presentation at the Workshop on Multiword Expressions: Identifying and Exploiting Underlying Properties. The workshop is endorsed by the Association for Computational Linguistics Special Interest Group on the Lexicon (SIGLEX) and is hosted in conjunction with the COLING/ACL 2006 on July 23rd, 2006 in Sydney, Australia. There has been a growing awareness in the NLP community of the problems that multiword expressions (MWEs) pose. Developments in areas such as machine translation, text summarization, paraphrasing, grammar development and parsing, information retrieval, and question answering (to mention a few) have acknowledged difficulties due to the idiosyncratic nature of multiword expressions. This workshop continues a tradition of ACL workshops on Collocations (2001) and Multiword Expressions (2003 and 2004). Its specific objective is to focus on the underlying properties of MWEs. The call for papers expressed our interest in several topics such as the definition of MWEs, properties of MWEs and their impact on NLP applications, representation and treatment of the different classes of MWEs, linguistic and psycholinguistic analyses of MWEs, evaluation of extraction techniques and the importance of (non-)compositionality. We received 23 submissions in total. Each submission was reviewed by (at least) three members of the program committee who not only judged each submission but also gave detailed comments to the authors. Among the received papers, 10 were selected for presentation at the workshop. After 3 papers have been withdrawn by their authors, seven papers are included in these proceedings. The intention of this workshop is to focus on some fundamental questions on the nature of MWEs. To do this we will allow plenty of time for discussion to pursue some of the interesting, open and difficult questions that MWEs raise. As well as a discussion period after each session of papers, we will be organising group discussions at the end of the workshop. These will focus on problems of defining, characterising and evaluating MWEs, given what we know about the range of phenomena that they encompass as well as any important questions that have arisen during the workshop.