Geometry of Compositionality

Hongyu Gong, Suma Bhat, Pramod Viswanath

AAAI 2017 pp. 3202-3208

doi:10.1609/AAAI.V31I1.10990 /aaai/2017/gong2017aaai-geometry/

Abstract

This paper proposes a simple test for compositionality (i.e., literal usage) of a word or phrase in a context-specific way. The test is computationally simple, relying on no external resources and only uses a set of trained word vectors. Experiments show that the proposed method is competitive with state of the art and displays high accuracy in context-specific compositionality detection of a variety of natural language phenomena (idiomaticity, sarcasm, metaphor) for different datasets in multiple languages. The key insight is to connect compositionality to a curious geometric property of word embeddings, which is of independent interest.

PDF AAAI Semantic Scholar

Cite

Text

Gong et al. "Geometry of Compositionality." AAAI Conference on Artificial Intelligence, 2017. doi:10.1609/AAAI.V31I1.10990

Markdown

[Gong et al. "Geometry of Compositionality." AAAI Conference on Artificial Intelligence, 2017.](https://mlanthology.org/aaai/2017/gong2017aaai-geometry/) doi:10.1609/AAAI.V31I1.10990

BibTeX

@inproceedings{gong2017aaai-geometry,
  title     = {{Geometry of Compositionality}},
  author    = {Gong, Hongyu and Bhat, Suma and Viswanath, Pramod},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2017},
  pages     = {3202-3208},
  doi       = {10.1609/AAAI.V31I1.10990},
  url       = {https://mlanthology.org/aaai/2017/gong2017aaai-geometry/}
}