01
Multilingual Text Style Transfer: Datasets & Models for Indian Languages
Proceedings of INLG 2024 — 17th International Natural Language Generation Conference2024
Postdoctoral Researcher
Sourabrata
Sourabrata
Mukherjee
Studying how language models understand, generate, and reason across languages and cultures.
About
I am a postdoctoral researcher at Microsoft Research, working on language models with a focus on multilingual and cultural understanding, controllable generation, and the evaluation of large models.
I recently completed my Ph.D. in Computational Linguistics at the Institute of Formal and Applied Linguistics, Charles University in Prague, where I was advised by Prof. Ondřej Dušek. My dissertation, Text Style Transfer using Neural Models, develops methods for rewriting text under attribute constraints — formality, sentiment, politeness, toxicity — across high- and low-resource languages.
Before the Ph.D., I spent over five years building production ML and analytics systems as a software and machine‑learning engineer in industry, and I held research roles at UKP Lab (TU Darmstadt), MBZUAI with Prof. Monojit Choudhury, and Panlingua. I hold an M.Tech from NIT Durgapur.
Research interests
What I work on
I build and study language technology that is useful across the world's languages — not only the few well‑resourced ones. My recent work spans:
News
Recent updates
- 2026 · now Joined Microsoft Research as a postdoctoral researcher, working on language AI & multilingual evaluation.
- 2025 Successfully defended my Ph.D. thesis “Text Style Transfer using Neural Models” at Charles University, Prague.
- 2024 · summer Research internship at MBZUAI (Abu Dhabi) on culture‑aware LLMs with Prof. Monojit Choudhury.
- 2024 · jul Two papers accepted at INLG 2024 — multilingual style transfer for Indian languages, and an evaluation of LLMs at style transfer.
- 2023 · dec Best Paper Award at the BLP Workshop, EMNLP 2023, for low‑resource Bangla style transfer.
- 2023 Awarded the Charles University Grant Agency (GAUK) research grant as PI; completed with honourable mention.
Selected publications
view all on google scholar
Papers
A complete and up‑to‑date list lives on Google Scholar. Below are selected works (* indicates equal contribution where applicable).
02
Are Large Language Models Actually Good at Text Style Transfer?
Proceedings of INLG 20242024
03
A Survey of Text Style Transfer: Applications and Ethical Implications
Northern European Journal of Language Technology (NEJLT)2024
04
Text Style Transfer: An Introductory Overview
4EU+ International Workshop on Recent Advancements in Artificial Intelligence2024
05
Low‑Resource Text Style Transfer for Bangla: Data & Models
Proceedings of the First Workshop on Bangla Language Processing (BLP), EMNLP 2023★ Best Paper2023
06
Text Detoxification as Style Transfer in English and Hindi
Proceedings of ICON 2023 — 20th International Conference on Natural Language Processing2023
07
Leveraging Low‑resource Parallel Data for Text Style Transfer
Proceedings of INLG 20232023
08
Polite Chatbot: A Text Style Transfer Application
Proceedings of EACL 2023 — Student Research Workshop2023
09
Balancing the Style‑Content Trade‑off in Sentiment Transfer using Polarity‑Aware Denoising
International Conference on Text, Speech, and Dialogue (TSD)2022
Experience
Where I've been
-
Postdoctoral Researcher — Microsoft Research
2026 — presentLanguage AI: multilingual & cultural NLP, LLM evaluation and LLM‑as‑a‑judge, controllable generation.
-
Ph.D. in Computational Linguistics — Charles University, Prague
2019 — 2025Institute of Formal and Applied Linguistics. Advisor: Prof. Ondřej Dušek. Thesis: Text Style Transfer using Neural Models.
-
Visiting Researcher (Internship) — MBZUAI, Abu Dhabi
Summer 2024Culture & LLM project with Prof. Monojit Choudhury.
-
Research Intern — Panlingua, India
2022 — 2023Low‑resource machine translation for Indian languages.
-
Research Assistant — UKP Lab, TU Darmstadt
2018 — 2019Built a data‑to‑text generation corpus by automatically detecting evidence linked to structured tables in scientific papers.
-
Technical Lead — Tricon Infotech, Bangalore
2016 — 2019Led an analytical recommendation engine driven by usage statistics.
-
Senior Software Engineer — Avaya Research Lab, Bangalore
2016Full‑stack log monitoring with optimised persistence.
-
Software Engineer — O9 Solutions, Bangalore
2015 — 2016Batch‑job management framework with notification capability.
-
Senior Software Engineer — Amdocs, Pune
2013 — 2015Recommendation engine and search components for a large web application.
-
M.Tech, Computer Science & Engineering — NIT Durgapur
2011 — 2013Advisor: Prof. Tandra Pal. Evolutionary algorithms for GFRP composites machining and high‑level synthesis.
Honours
Awards & grants
Best Paper Award — BLP Workshop, EMNLP 2023
For Low‑Resource Text Style Transfer for Bangla.
GAUK Research Grant (PI)
Charles University Grant Agency grant; project completed with honourable mention.
GATE Government Scholarship
Awarded for the Graduate Aptitude Test in Engineering during M.Tech at NIT Durgapur.
IBM National Technical Competition — Winner
National‑level technical competition organised by IBM.
Selected talks
Talks & presentations
Text Style Transfer: An Introductory Overview
Invited overview at the 4EU+ International Workshop on Recent Advancements in AI.
Two papers presented at INLG 2024
Multilingual style transfer for Indian languages; LLMs at text style transfer.
Low‑Resource Text Style Transfer for Bangla
Best paper presentation at the Bangla Language Processing workshop.
Polite Chatbot — Style Transfer in Dialogue
Student Research Workshop presentation at EACL 2023.
Teaching
Teaching & mentoring
I have taught and developed introductory material on Machine Learning, Deep Learning, Programming, Data Structures, and Search Engine Optimization. During my Ph.D. I co‑mentored junior researchers and interns on text generation projects.
Ph.D. mentoring & collaboration
Worked alongside master's and Ph.D. students on style transfer, detoxification, and politeness‑aware dialogue.
Introductory courses (ML / DL / DS)
Designed and delivered short courses on ML/DL, programming, and data structures for students and early‑career engineers.
Writing
Notes & essays
A growing space for shorter‑form writing on language models, evaluation, and what it means for AI to work across cultures.
First posts coming soon — drafts are in progress on multilingual evaluation and what culture means inside an LLM. Check back, or subscribe via my Scholar profile.