GEMS -- GEometrical Models of natural language Semantics Workshop at EMNLP, July 31st 2011, Edinburgh, Scotland http://sites.google.com/site/geometricalmodels/
GEMS 2011 is the third event in a series of workshops on distributional models, also known as semantic spaces. These models have become omnipresent in computational linguistics and neighboring fields.
GEMS 2011 invites original contributions to problems in meaning representation, acquisition and use, based on distributional and vector space models. We are interested in methodological innovations as well as tasks ranging from the induction of linguistic and world knowledge to practical and industrial NLP applications.
GEMS 2011 will also address one particular challenge of geometrical models as a scientific field, namely fragmentation -- with respect to data sets, methods and evaluation metrics. To facilitate the comparison of studies and achieve scientific progress, GEMS will introduce a shared evaluation:
- We provide two datasets suitable for the evaluation of distributional models through our website, together with the corpora that can be used for their modeling. - These datasets cover two major tasks: differentiating between semantic relations and addressing compositionality. - Papers submitted to GEMS are strongly encouraged to evaluate their models on one of the datasets, or, if this is not possible, to discuss why their models are not applicable.
Topics of interest include, but are not limited to
- Document-based, collocational and syntax-based spaces - Eigenvector methods and geometrical embeddings - Higher order tensors - Computational complexity and evaluation issues - Graph-based models over semantic spaces - Logic and inference in semantic spaces - Large-scale implementations of distributional models (Map-Reduce, Hadoop) - Kernels methods for NLP - Word sense discrimination and discrimination; lexical substitution - Induction of selectional preferences - Compositionality in geometrical models: phrase representation; concept combination - Lexicon acquisition - Conceptual clustering - Modeling of linguistic theories and ontological knowledge - Cognitive theories of semantic space models - Applications in the humanities and social sciences - Applications and impact on Web search, Web mining, Query log mining, Query Intent Modeling, and other industrial activities
Authors are invited to submit papers on original, unpublished work on the topics of this workshop. There are three paper categories: long papers, short papers and demos.
- Long papers should present completed work. They can have up to 9 pages of content, plus references. - Short papers/demos can present work in progress or the description of a system. They must not exceed 4 pages plus one page of references.
Submissions should follow the two-column format of the ACL 2011 proceedings (see the official style files at http://www.acl2011.org/call.shtml). As reviewing will be blind, please ensure that papers are anonymous. The papers should not include the authors' names and affiliations or any references to web sites, project names, etc. revealing the authors' identity.
Each submission will be reviewed by at least two members of the program committee. Accepted papers will be published in the workshop proceedings.
Papers should be submitted electronically at https://www.softconf.com/emnlp/GEMS2011/
April 22, 2011: Papers due May 20, 2011: Notification of acceptance June 03, 2011: Camera-ready deadline July 31, 2011: Workshop
Sebastian Pado, University of Heidelberg (Chair) Yves Peirsman, Stanford University & KULeuven (Chair)