On the Role of Pre-trained Language Models in Word Ordering: A Case Study with BART
This work addresses word ordering for natural language processing, but it is incremental as it applies an existing model to a new task.
The study tackled the problem of word ordering using pre-trained language models, specifically BART, and found it effective, with performance gains also reported in the related partial tree linearization task.
Word ordering is a constrained language generation task taking unordered words as input. Existing work uses linear models and neural networks for the task, yet pre-trained language models have not been studied in word ordering, let alone why they help. We use BART as an instance and show its effectiveness in the task. To explain why BART helps word ordering, we extend analysis with probing and empirically identify that syntactic dependency knowledge in BART is a reliable explanation. We also report performance gains with BART in the related partial tree linearization task, which readily extends our analysis.