Fast Rule-Based Decoding: Revisiting Syntactic Rules in Neural Constituency Parsing
This work addresses the bottleneck of decoder inefficiency in neural constituency parsing for NLP researchers, offering a novel method that enhances parsing accuracy and cross-domain applicability.
The paper tackles the problem of neural constituency parsing by introducing a fast, GPU-accelerated CKY decoding procedure that incorporates syntactic rules, achieving 95.89 and 92.52 F1 scores on PTB and CTB datasets, respectively, with significant improvements over prior methods and strong cross-domain performance in zero-shot settings.
Most recent studies on neural constituency parsing focus on encoder structures, while few developments are devoted to decoders. Previous research has demonstrated that probabilistic statistical methods based on syntactic rules are particularly effective in constituency parsing, whereas syntactic rules are not used during the training of neural models in prior work probably due to their enormous computation requirements. In this paper, we first implement a fast CKY decoding procedure harnessing GPU acceleration, based on which we further derive a syntactic rule-based (rule-constrained) CKY decoding. In the experiments, our method obtains 95.89 and 92.52 F1 on the datasets of PTB and CTB respectively, which shows significant improvements compared with previous approaches. Besides, our parser achieves strong and competitive cross-domain performance in zero-shot settings.