Lieve Macken is a computational linguist with a long track record in language technology. She started her career at the Center for Computational Linguistics of the University of Leuven, developing a Simplified English Grammar Checker and Corrector (SECC). The grammar and style checker was intended as a front-end for machine translation systems helping to improve translation quality and reduce post-editing work by simplifying the input.
She worked at the R&D-department of Lernout & Hauspie Speech products and Scansoft for more than 8 years. She developed, first as language specialist, and later as a project and team leader text-to-speech systems for English.
From 2006 to 2009, she had the technical leadership of the Dutch Parallel Corpus project. In 2010 she defended her PhD "Sub-sentential alignment of translational correspondences", which formed the basis of the TExSIS bilingual terminology extraction system.
Her research focuses on the impact of translation technology and Machine Translation on the process and product of translation.
As promotor of the ROBOT project she supervised the PhD of Joke Daems. In the ROBOT project the translation behavior of professional and novice translators has been observed using keystroke logging and eye-tracking software. In the framework of the SCATE project, she supervised the PhD of Arda Tezcan on Quality Estimation of Machine Translation, with a focus on estimating post-editing effort.
She is currently promotor of two ongoing projects. ArisToCAT, Assessing The Comprehensibility of Automatic Translations (FWO) and PreDicT, Predicting Difficulty in Translation (BOF UGent).
She is the operational head of the language technology section of the department, where she also teaches Translation Technology, Machine Translation, and Localisation.