Information in natural language is almost always located in time and/or space. The coherence of a document also results from a coherent temporal perspective, and in certain cases from the coherence of the spatial frame.
Beyond specific applications where this information is obviously central (navigation systems, scene visualization, geo-localization, chronology extraction, etc), time and space also play an important part in other related tasks, such as information extraction where they constrain the validity of most facts, or in summarization and question-answering where temporal coherence is important. From a more fundamental perspective, temporal and spatial modifications have a pervasive influence in semantic and pragmatic representations. It is natural that they motivate a growing number of projects, specific annotated corpora, which in turn lead to specification and normalization efforts, such as the ISO-TimeML and ISO-Space standards.
All this have helped to move from essentially theoretical models in the 1990s, to arrive at a body of work with better empirically justified foundations. This can be shown with recent evaluation campaigns for some sub-tasks (Tempeval) within the Semeval campaigns.
Expressions of time and space also have some specific aspects that set them apart from other semantic tasks in natural language processing. A notable aspect is the importance of the underlying semantics of temporal and spatial predicates, as is noted in several studies that take their inferential properties into account, especially when considered at the level of a document. Evaluation of the resulting representations is then not a trivial matter.
The objective of this special issue is to present new developments in the processing of temporal and spatial information in language, from theoretical, practical and methodological point of views. Spatial processing has been of increasing interest lately, and raises specific issues, and we encourage work that focus on this aspect, in isolation or in relation with temporal information. Presentation of the importance of such information in applications is also encouraged.
We encourage submission on any aspect related to the processing of temporal and spatial information in natural language, especially on the following issues and tasks:
temporal information extraction : date and event extraction, event anchoring, relating
events within a text or a text collection
spatial information extraction: spatial entity recognition, geolocalisation, relating spatial entities
joint processing of temporal and spatial information, motion description, route description
representation and reasoning issues, in particular interoperability of temporal/spatial representations
generation of scene and image description
information extraction and monitoring in specific domains, such as the medical domain
joint processing of time and modality
annotation scheme for time and space
creation and use of resources for temporal and spatial processing
specificities of evaluation procedures for temporal/spatial annotation schemes, resources and processes
Robert J. Bobrow
Deadline for submission: March 15th, 2012
List of papers selected: July 2012
Deadline for camera ready papers: September 2012
Publication on line: end of 2012
TAL (Traitement Automatique des Langues / Natural Language Processing) is a forty year old international journal published by ATALA (French Association for Natural Language Processing) with the support of CNRS (National Centre for Scientific Research). It has moved to an electronic mode of publication, with printing on demand. This affects in no way its reviewing and selection process.
Authors intending to submit a paper are encouraged to contact the guest editors of the issue: Inderjeet Mani (inderjeet.mani at gmail.com) and Philippe Muller (philippe.muller at irit.fr)
Contributions (around 25 pages, PDF format) must be uploaded at http://tal-53-2.sciencesconf.org/
Style sheets are available for download on the Web site of the journal.
The journal only publishes original contributions in French or in English.