About me

ආයුබෝවන් (āyubōvan), May you live longer!!

I am a PhD Student at Research Group in Computational Linguistics (RGCL) at the University of Wolverhampton supervised by Prof Ruslan Mitkov and Prof Constantin Orasan. My current research focuses on using Deep Learning for the applications in translation technology.

I received my bachelor’s degree from the University of Moratuwa, Sri Lanka, specialising in Computer Science & Engineering. After graduation, I worked as an Associate Technical Team Lead in the R&D division of CodeGen International where I lead and managed the Machine Learning and Data Science team. I started my PhD in 2018.

I am interested in Deep Learning Applications in Natural Language Processing.

Recent News

  • On 5th May 2021, our paper on translation quality estimation - “An Exploratory Analysis of Multilingual Word Level Quality Estimation with Cross-Lingual Transformers” was accepted to ACL-IJCNLP 2021. We extend the TransQuest framework for word level quality estimation.

  • On March 18th 2020, Our paper on offensive spans detection - “MUDES - {Mu}ltilingual {De}tection of Offensive {S}pans” was accepted to NAACL 2021 as a system demonstration paper.

  • On September 30th 2020, Our work on Translation Quality Estimation - “TransQuest: Translation Quality Estimation with Cross-lingual Transformers” was accepted to COLING 2020 as a long paper. We release a QE framework for sentence-level quality estimation.

  • On September 15th 2020, a short paper accepted to EMNLP 2020: “Multilingual Offensive Language Identification with Cross-lingual Embeddings” - We present our initial experiments on transfer learning for Offensive Language Identification in low resource languages.

  • In August 2020, Our framework developed for Quality Estimation: TransQuest won the first place in WMT 2020 sentence-level DA shared task in all the language pairs and the multilingual track out of 50 participants.

  • On April 21st 2020, A paper titled “Intelligent Translation Memory Matching and Retrieval with Sentence Encoders” was accepted in EAMT 2020. We release our early findings on using deep learning for translation memory matching and retrieval.

  • On February 15th 2020. A paper titled “Offensive Language Identification in Greek” was accepted in LREC 2020. We release an Offensive Language Identification dataset for Greek as well as several baseline models to tackle the task.