Improving Sparse Word Representations with Distributional Inference for Semantic Composition
TL
TL;DR
This paper shows that distributional inference improves sparse word representations on several word similarity benchmarks and demonstrates that the model is competitive with the state-of-the-art for adjective-noun, noun- noun and verb-object compositions while being fully interpretable.
Abstract
Distributional models are derived from co-occurrences in a corpus, where only a small proportion of all possible plausible co-occurrences will be observed. This results in a very sparse vector space, requiring a mechanism for inferring missing knowledge. Most methods face this challenge in ways that render the resulting word representations uninterpretable, with the consequence that semantic composition becomes hard to model. In this paper we explore an alternative which involves explicitly inferring unobserved co-occurrences using the distributional neighbourhood. We show that distributional inference improves sparse word representations on several word similarity benchmarks and demonstrate that our model is competitive with the state-of-the-art for adjective-noun, noun-noun and verb-object compositions while being fully interpretable.
References55 items
1
Aligning Packed Dependency Trees: A Theory of Composition for Distributional Semantics
2
The Forest Convolutional Network: Compositional Distributional Semantics with a Neural Chart and without Binarization
3
From Paraphrase Database to Compositional Paraphrase Model and Back
4
Improving Distributional Similarity with Lessons Learned from Word Embeddings
5