Affiliation:
1. Katholieke Universiteit Leuven, Belgium
Abstract
This article reports on the novel task of
spatial role labeling
in natural language text. It proposes machine learning methods to extract spatial roles and their relations. This work experiments with both a step-wise approach, where spatial prepositions are found and the related trajectors, and landmarks are then extracted, and a joint learning approach, where a spatial relation and its composing indicator, trajector, and landmark are classified collectively. Context-dependent learning techniques, such as a skip-chain conditional random field, yield good results on the GUM-evaluation (Maptask) data and CLEF-IAPR TC-12 Image Benchmark. An extensive error analysis, including feature assessment, and a cross-domain evaluation pinpoint the main bottlenecks and avenues for future research.
Publisher
Association for Computing Machinery (ACM)
Subject
Computational Mathematics,Computer Science (miscellaneous)
Reference49 articles.
1. Prepositions in Applications: A Survey and Introduction to the Special Issue
2. Barclay M.
and
Galton A
.
2008
. An influence model for reference object selection in spatially locative phrases. In Spatial Cognition VI: Learning Reasoning and Talking about Space C. Freksa N. S. Newcombe P. Gärdenfors and S. Wölfl Eds. Lecture Notes in Artificial Intelligence vol.
5241 Springer 216--232. 10.1007/978-3-540-87601-4_17 Barclay M. and Galton A. 2008. An influence model for reference object selection in spatially locative phrases. In Spatial Cognition VI: Learning Reasoning and Talking about Space C. Freksa N. S. Newcombe P. Gärdenfors and S. Wölfl Eds. Lecture Notes in Artificial Intelligence vol. 5241 Springer 216--232. 10.1007/978-3-540-87601-4_17
3. The Role of Conceptual and Linguistic Ontologies in Interpreting Spatial Discourse
4. Language and Space: a two-level semantic approach based on principles of ontological engineering
Cited by
53 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献