Token-Level Supervised Contrastive Learning for Punctuation Restoration
This work addresses punctuation restoration for improving downstream NLP tasks like intent detection and slot filling, but it is incremental as it builds on existing pre-trained language models.
The paper tackles punctuation restoration in automatic speech recognition by proposing a token-level supervised contrastive learning method to address data imbalance, resulting in up to 3.2% absolute F1 improvement on the test set.
Punctuation is critical in understanding natural language text. Currently, most automatic speech recognition (ASR) systems do not generate punctuation, which affects the performance of downstream tasks, such as intent detection and slot filling. This gives rise to the need for punctuation restoration. Recent work in punctuation restoration heavily utilizes pre-trained language models without considering data imbalance when predicting punctuation classes. In this work, we address this problem by proposing a token-level supervised contrastive learning method that aims at maximizing the distance of representation of different punctuation marks in the embedding space. The result shows that training with token-level supervised contrastive learning obtains up to 3.2% absolute F1 improvement on the test set.