Abstract
This paper describes the lexical simplification system RCML submitted to the English language track of the TSAR-2022 Shared Task. The system leverages a pre-trained language model to generate contextually plausible substitution candidates which are then ranked according to their simplicity as well as their grammatical and semantic similarity to the target complex word. Our submissions secure 6th and 7th places out of 33, improving over the SOTA baseline for 27 out of the 51 metrics.- Anthology ID:
- 2022.tsar-1.29
- Volume:
- Proceedings of the Workshop on Text Simplification, Accessibility, and Readability (TSAR-2022)
- Month:
- December
- Year:
- 2022
- Address:
- Abu Dhabi, United Arab Emirates (Virtual)
- Venue:
- TSAR
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 259–263
- Language:
- URL:
- https://aclanthology.org/2022.tsar-1.29
- DOI:
- 10.18653/v1/2022.tsar-1.29
- Cite (ACL):
- Desislava Aleksandrova and Olivier Brochu Dufour. 2022. RCML at TSAR-2022 Shared Task: Lexical Simplification With Modular Substitution Candidate Ranking. In Proceedings of the Workshop on Text Simplification, Accessibility, and Readability (TSAR-2022), pages 259–263, Abu Dhabi, United Arab Emirates (Virtual). Association for Computational Linguistics.
- Cite (Informal):
- RCML at TSAR-2022 Shared Task: Lexical Simplification With Modular Substitution Candidate Ranking (Aleksandrova & Brochu Dufour, TSAR 2022)
- PDF:
- https://preview.aclanthology.org/remove-xml-comments/2022.tsar-1.29.pdf