ATLIS: Identifying Locational Information in Text Automatically

John Vogel, Marc Verhagen, James Pustejovsky


Abstract
ATLIS (short for “ ATLIS Tags Locations in Strings”) is a tool being developed using a maximum-entropy machine learning model for automatically identifying information relating to spatial and locational information in natural language text. It is being developed in parallel with the ISO-Space standard for annotation of spatial information (Pustejovsky, Moszkowicz & Verhagen 2011). The goal of ATLIS is to be able to take in a document as raw text and mark it up with ISO-Space annotation data, so that another program could use the information in a standardized format to reason about the semantics of the spatial information in the document. The tool (as well as ISO-Space itself) is still in the early stages of development. At present it implements a subset of the proposed ISO-Space annotation standard: it identifies expressions that refer to specific places, as well as identifying prepositional constructions that indicate a spatial relationship between two objects. In this paper, the structure of the ATLIS tool is presented, along with preliminary evaluations of its performance.
Anthology ID:
L12-1610
Volume:
Proceedings of the Eighth International Conference on Language Resources and Evaluation (LREC'12)
Month:
May
Year:
2012
Address:
Istanbul, Turkey
Venue:
LREC
SIG:
Publisher:
European Language Resources Association (ELRA)
Note:
Pages:
612–616
Language:
URL:
http://www.lrec-conf.org/proceedings/lrec2012/pdf/1022_Paper.pdf
DOI:
Bibkey:
Cite (ACL):
John Vogel, Marc Verhagen, and James Pustejovsky. 2012. ATLIS: Identifying Locational Information in Text Automatically. In Proceedings of the Eighth International Conference on Language Resources and Evaluation (LREC'12), pages 612–616, Istanbul, Turkey. European Language Resources Association (ELRA).
Cite (Informal):
ATLIS: Identifying Locational Information in Text Automatically (Vogel et al., LREC 2012)
Copy Citation:
PDF:
http://www.lrec-conf.org/proceedings/lrec2012/pdf/1022_Paper.pdf