Score Combination for Improved Parallel Corpus Filtering for Low Resource Conditions
Muhammad ElNokrashy, Amr Hendy, Mohamed Abdelghaffar, Mohamed Afify, Ahmed Tawfik, Hany Hassan Awadalla
Abstract
This paper presents the description of our submission to WMT20 sentence filtering task. We combine scores from custom LASER built for each source language, a classifier built to distinguish positive and negative pairs and the original scores provided with the task. For the mBART setup, provided by the organizers, our method shows 7% and 5% relative improvement, over the baseline, in sacreBLEU score on the test set for Pashto and Khmer respectively.- Anthology ID:
- 2020.wmt-1.106
- Volume:
- Proceedings of the Fifth Conference on Machine Translation
- Month:
- November
- Year:
- 2020
- Address:
- Online
- Venue:
- WMT
- SIG:
- SIGMT
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 947–951
- Language:
- URL:
- https://aclanthology.org/2020.wmt-1.106
- DOI:
- Cite (ACL):
- Muhammad ElNokrashy, Amr Hendy, Mohamed Abdelghaffar, Mohamed Afify, Ahmed Tawfik, and Hany Hassan Awadalla. 2020. Score Combination for Improved Parallel Corpus Filtering for Low Resource Conditions. In Proceedings of the Fifth Conference on Machine Translation, pages 947–951, Online. Association for Computational Linguistics.
- Cite (Informal):
- Score Combination for Improved Parallel Corpus Filtering for Low Resource Conditions (ElNokrashy et al., WMT 2020)
- PDF:
- https://preview.aclanthology.org/nodalida-main-page/2020.wmt-1.106.pdf