Research Area: Machine Learning
Distributional semantics based on neural approaches is a cornerstone of Natural Language Processing, with surprising connections to human meaning representation as well. Recent Transformer-based Language Models have proven capable of producing contextual word representations that reliably convey sense-specific information, simply as a product of self-supervision. Prior work has shown that these contextual representations can be used to accurately represent large sense inventories as sense embeddings, to the extent that a distance-based solution to Word Sense Disambiguation (WSD) tasks outperforms models trained specifically for the task. Still, there remains much to understand on how to use these Neural Language Models (NLMs) to produce sense embeddings that can better harness each NLMs meaning representation abilities. In this work we introduce a more principled approach to leverage information from all layers of NLMs, informed by a probing analysis on 14 NLM variants. We also emphasize the versatility of these sense embeddings in contrast to task-specific models, applying them on several sense-related tasks, besides WSD, while demonstrating improved performance using our proposed approach over prior work focused on sense embeddings. Finally, we discuss unexpected findings regarding layer and model performance variations, and potential applications for downstream tasks.
Keywords:
LMMS reloaded
Transformer
Word Sense Disambiguation (WSD)
Neural Language Models
Deep Learning
Machine Learning
Author(s) Name: Daniel Loureiro, Alípio Mário Jorge, Jose Camacho-Collados
Journal name: Artificial Intelligence
Conferrence name:
Publisher name: Elsevier
DOI: 10.1016/j.artint.2022.103661
Volume Information: Volume 305, April 2022, 103661
