Addressing Syntax-Based Semantic Complementation: Incorporating Entity and Soft Dependency Constraints into Metonymy Resolution

Abstract

State-of-the-art methods for metonymy resolution (MR) consider the sentential context by modeling the entire sentence. However, entity representation, or syntactic structure that are informative may be beneficial for identifying metonymy. Other approaches only using deep neural network fail to capture such information. To leverage both entity and syntax constraints, this paper proposes a robust model EBAGCN for metonymy resolution. First, this work extracts syntactic dependency relations under the guidance of syntactic knowledge. Then the work constructs a neural network to incorporate both entity representation and syntactic structure into better resolution representations. In this way, the proposed model alleviates the impact of noisy information from entire sentences and breaks the limit of performance on the complicated texts. Experiments on the SemEval and ReLocaR dataset show that the proposed model significantly outperforms the state-of-the-art method BERT by more than 4%. Ablation tests demonstrate that leveraging these two types of constraints benefits fine pre-trained language models in the MR task.

Feature Selection for Location Metonymy Using Augmented Bag-of-Words

“…The results of their experiments achieve the state of the art with 89.1%, 94.8% and 95.9% in SemEval, ReLocaR and WIMCOR respectively. Du and Wang [41] state that the state-of-the-art methods only focus on the context of the sentence neglecting Entity representation and Syntactic structure, their study is fixated on exploiting entity and syntax constraints by obtaining syntactic dependency relations, then developing a neural network that can integrate both constraints for a better representation, thus, enabling the model's performance to surpass the limitations when encountering complex sentences along with reducing the noise. The experiments on SemEval and ReLocaR datasets display a considerable improvement where it scored 89.8% and 95.7% respectively which is over 4% compared to the BERT model [22].…”