A Novel Approach for Root Selection in the Dependency Parsing
Sharefah Ahmed Al-Ghamdi, Hend Al-Khalifa, Abdulmalik AlSalman
Abstract
Although syntactic analysis using the sequence labeling method is promising, it can be problematic when the labels sequence does not contain a root label. This can result in errors in the final parse tree when the postprocessing method assumes the first word as the root. In this paper, we present a novel postprocessing method for BERT-based dependency parsing as sequence labeling. Our method leverages the root’s part of speech tag to select a more suitable root for the dependency tree, instead of using the default first token. We conducted experiments on nine dependency treebanks from different languages and domains, and demonstrated that our technique consistently improves the labeled attachment score (LAS) on most of them.- Anthology ID:
- 2024.osact-1.5
- Volume:
- Proceedings of the 6th Workshop on Open-Source Arabic Corpora and Processing Tools (OSACT) with Shared Tasks on Arabic LLMs Hallucination and Dialect to MSA Machine Translation @ LREC-COLING 2024
- Month:
- May
- Year:
- 2024
- Address:
- Torino, Italia
- Editors:
- Hend Al-Khalifa, Kareem Darwish, Hamdy Mubarak, Mona Ali, Tamer Elsayed
- Venues:
- OSACT | WS
- SIG:
- Publisher:
- ELRA and ICCL
- Note:
- Pages:
- 46–49
- Language:
- URL:
- https://aclanthology.org/2024.osact-1.5
- DOI:
- Cite (ACL):
- Sharefah Ahmed Al-Ghamdi, Hend Al-Khalifa, and Abdulmalik AlSalman. 2024. A Novel Approach for Root Selection in the Dependency Parsing. In Proceedings of the 6th Workshop on Open-Source Arabic Corpora and Processing Tools (OSACT) with Shared Tasks on Arabic LLMs Hallucination and Dialect to MSA Machine Translation @ LREC-COLING 2024, pages 46–49, Torino, Italia. ELRA and ICCL.
- Cite (Informal):
- A Novel Approach for Root Selection in the Dependency Parsing (Al-Ghamdi et al., OSACT-WS 2024)
- PDF:
- https://preview.aclanthology.org/nschneid-patch-5/2024.osact-1.5.pdf