Machine Translation (MT) for low-resource languages remains far behind that of high-resource languages, and the gap is widest in specialised domains, where parallel data is scarce or entirely absent. We present EnSiTa, a trilingual multi-domain parallel dataset and benchmark for English, Sinhala and Tamil. EnSiTa provides human post-edited training data for seven domains, plus manually translated test sets for those and one additional domain, all produced by professional translators under a multiyear, rigorously quality-controlled process. Using this dataset, we conduct an extensive study of domain-specific MT for all six language directions, fine-tuning a from-scratch Transformer, a pre-trained translation model (NLLB-600M), and decoder-only LLMs (Gemma 3 family, 1BâÂÂ12B, and TranslateGemma) across training-data sizes, model scales, and in-domain, cross-domain, multilingual and multi-domain settings. To the best of our knowledge, this is the most extensive systematically documented multi-domain parallel data creation and benchmarking effort for low-resource MT. Our data and models will be publicly released.
Keywords: Natural Language Processing | Sinhala |
@article{ranathungaen2026sita,
title={{EnSiTa-A Trilingual Multi-Domain Parallel Dataset and Benchmark for Domain-Specific Machine Translation}},
author={Ranathunga, Surangika and de Silva, Nisansa and Fernando, Aloka and Warnakulasuriya, Kavindu and Wijesiri, Isuru and Velayuthan, Menan and Rathnayaka, Charitha and Varatharajan, Thivaharan and Silva, Sajeevi and Kandanaarachchi, Piumi and Thayasivam, Uthayasanker},
journal={ResearchGate},
year={2026},
abstract={Machine Translation (MT) for low-resource languages remains far behind that of high-resource languages, and the gap is widest in specialised domains, where parallel data is scarce or entirely absent. We present EnSiTa, a trilingual multi-domain parallel dataset and benchmark for English, Sinhala and Tamil. EnSiTa provides human post-edited training data for seven domains, plus manually translated test sets for those and one additional domain, all produced by professional translators under a multiyear, rigorously quality-controlled process. Using this dataset, we conduct an extensive study of domain-specific MT for all six language directions, fine-tuning a from-scratch Transformer, a pre-trained translation model (NLLB-600M), and decoder-only LLMs (Gemma 3 family, 1BâÂÂ12B, and TranslateGemma) across training-data sizes, model scales, and in-domain, cross-domain, multilingual and multi-domain settings. To the best of our knowledge, this is the most extensive systematically documented multi-domain parallel data creation and benchmarking effort for low-resource MT. Our data and models will be publicly released.},
misc={https://www.researchgate.net/profile/Surangika-Ranathunga/publication/413704621_EnSiTa_-A_Trilingual_Multi-Domain_Parallel_Dataset_and_Benchmark_for_Domain-Specific_Machine_Translation/links/6a911c06d847465f626a8ea4/EnSiTa-A-Trilingual-Multi-Domain-Parallel-Dataset-and-Benchmark-for-Domain-Specific-Machine-Translation.pdf,NLP:SIN,https://goo.gl/iY6aTr}
}
ActivePI: Surangika Ranathunga
We propose to create a multi-domain Neural Machine Translation (NMT) System for Sinhala, Tamil, and English, the official languages of Sri Lanka.
ActivePI: Nisansa de Silva
We propose to advance natural language processing for low-resource settings by leveraging data augmentation, transfer learning, and model adaptation techniques to overcome the scarcity of annotated data and linguistic resources.
ActivePI: Nisansa de Silva
We propose to build and analyze Sinhala language resources, datasets, and models to strengthen NLP research for this low-resource language, while connecting it to English and Tamil through bilingual lexicons, embeddings, and cross-lingual methods.