Multi-sense Definition Modeling using Word Sense Decompositions
Preprint
Abstract
Word embeddings capture syntactic and semantic information about words. Definition modeling aims to make the semantic content in each embedding explicit, by outputting a natural language definition based on the embedding. However, existing definition models are limited in their ability to generate accurate definitions for different senses of the same word. In this paper, we introduce a new method that enables definition modeling for multiple senses. We show how a Gumble-Softmax approach outperforms baselines at matching sense-specific embeddings to definitions during training. In experiments, our multi-sense definition model improves recall over a state-of-the-art single-sense definition model by a factor of three, without harming precision.Keywords
All Related Versions
This publication has 8 references indexed in Scilit:
- Incorporating Sememes into Chinese Definition ModelingIEEE/ACM Transactions on Audio, Speech, and Language Processing, 2020
- Linear Algebraic Structure of Word Senses, with Applications to PolysemyTransactions of the Association for Computational Linguistics, 2018
- Conditional Generators of Words Definitions2018
- Deep Visual-Semantic Alignments for Generating Image DescriptionsIEEE Transactions on Pattern Analysis and Machine Intelligence, 2016
- Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation2014
- Efficient Non-parametric Estimation of Multiple Embeddings per Word in Vector Space2014
- WordNet: An electronic lexical database . Ed. by Christiane Fellbaum. Cambridge, MA: MIT Press, 1998. Pp. xxii, 423.Language, 2000
- Semantically significant patterns in dictionary definitions1986