Paper: Distributed Representations of Geographically Situated Language

ACL ID P14-2134
Title Distributed Representations of Geographically Situated Language
Venue Annual Meeting of the Association of Computational Linguistics
Session Main Conference
Year 2014

We introduce a model for incorporating contextual information (such as geogra- phy) in learning vector-space representa- tions of situated language. In contrast to approaches to multimodal representation learning that have used properties of the object being described (such as its color), our model includes information about the subject (i.e., the speaker), allowing us to learn the contours of a word?s meaning that are shaped by the context in which it is uttered. In a quantitative evaluation on the task of judging geographically in- formed semantic similarity between repre- sentations learned from 1.1 billion words of geo-located tweets, our joint model out- performs comparable independent models that learn meaning in isolation.