WAFFLE: A Graph for WordNet Applied to FreeForm Linguistic Exploration

EMNLP (NLPOSS) 2020  ·  Berk Ekmekci, Blake Howald ·

The WordNet database of English (Fellbaum, 1998) is a key source of semantic information for research and development of natural language processing applications. As the sophistication of these applications increases with the use of large datasets, deep learning, and graph-based methods, so should the use of WordNet. To this end, we introduce WAFFLE: WordNet Applied to FreeForm Linguistic Exploration which makes WordNet available in an open source graph data structure. The WAFFLE graph relies on platform agnostic formats for robust interrogation and flexibility. Where existing implementations of WordNet offer dictionary-like lookup, single degree neighborhood operations, and path based similarity-scoring, the WAFFLE graph makes all nodes (semantic relation sets) and relationships queryable at scale, enabling local and global analysis of all relationships without the need for custom code. We demonstrate WAFFLE’s ease of use, visualization capabilities, and scalable efficiency with common queries, operations, and interactions. WAFFLE is available at github.com/TRSS-NLP/WAFFLE.

PDF Abstract

Datasets


  Add Datasets introduced or used in this paper

Results from the Paper


  Submit results from this paper to get state-of-the-art GitHub badges and help the community compare results to other papers.

Methods


No methods listed for this paper. Add relevant methods here