I am a post-doctoral researcher in statistical machine translation and VENI laureate since September 2016. I work with a team of six PhD students led by Christof Monz, and I co-supervise three: Marlies van der Wees, Ke Tran and Marzieh Fadaee. Currently, the main goal of my work is to improve SMT into morphologically rich languages, for instance using class-based language models or neural translation models.
Interested in my work ? See my Research page.
- [Oct 2016] I'm coming back to work (very gradually) after an extraordinary summer spent on maternity leave!
- [Sep 2016] Paper accepted at COLING 2016: "Measuring the Effect of Conversational Aspects on Machine Translation Quality", M. van der Wees, A. Bisazza and C. Monz.
- [Jul 2016] I won a VENI grant (250K personal grant from the Dutch research funding agency) to fund my project "Connecting the Dots: Minimal Structure Modeling for Machine Translation"!
- [Jul 2016] Paper accepted at EMNLP 2016: "Neural versus Phrase-Based Machine Translation Quality: a Case Study", L. Bentivogli, A. Bisazza, M. Cettolo and M. Federico.
- [Jun 2016] My survey finally appears on Computational Linguistics: "A Survey of Word Reordering in Statistical Machine Translation: Computational Models and Language Phenomena", A.Bisazza and M. Federico.
- [Mar 2016] Paper accepted at NAACL 2016: "Recurrent Memory Networks for Language Modeling", K. Tran, A. Bisazza and C. Monz.
- [Feb 2016] We got a Google Research Award to partially fund my research on decoding-time generation for SMT into morphologically rich languages!
Research interests(for more details see my Research page)
- Translation from/into morphologically rich languages
- Neural language and translation modeling
- Enhancing statistical models with linguistically motivated techniques
- Word reordering between distant languages
- Domain adaptation
- Linguistic pre-processing and language modeling methods for morphologically rich languages
- Italian, English and French
- ... but more interestingly: Turkish (advanced speaking and reading) and Arabic (mostly reading).