Speculation and Negation Scope Detection via Convolutional Neural Networks

Zhong Qian1, Peifeng Li1, Qiaoming Zhu1, Guodong Zhou1, Zhunchen Luo2, Wei Luo2
1Soochow University, 2China Defense Science and Technology Information Center


Abstract

Speculation and negation are important in-formation to identify text factuality. In this paper, we propose a Convolutional Neural Network (CNN)-based model with probabil-istic weighted average pooling to address speculation and negation scope detection. In particular, our CNN-based model extracts those meaningful features from various syn-tactic paths between the cues and the candi-date tokens in both constituency and depen-dency parse trees. Evaluation on BioScope shows that our CNN-based model significant-ly outperforms the state-of-the-art systems on Abstracts, a sub-corpus in BioScope, and achieves comparable performances on Clini-cal Records, another sub-corpus in BioScope.