My research interest evolved from morphology and tagging of inflective languages (lexicons, analysis and generation tools - now reimplemented by Milan Straka as MorphoDiTa) to machine translation (French-English while at IBM and Czech-English; also, Czech-Russian and other closely related languages). I am also interested in parsing (see e.g. the CLSP Workshop on parsing Czech) and generation. However, in the past 10 years, I devoted most of my research time to creating linguistic resources, such as the Prague Dependency Treebank family of projects (Czech, English, Arabic) and managing new research projects, mainly funded by the EU (see below for a complete list). I am also involved in the Universal Dependencies project, led by Joakim Nivre of Uppsala University and hosted by LINDAT/CLARIAH-CZ as the official UD repository. My management responsibilities include the LINDAT/CLARIAH-CZ Research Infrastructure and also being the Deputy Chair of the Institute.
I am also interested in spoken language understanding. I participated in the now finishing project Malach, both on the language modeling part (for ASR), on thesaurus translation and on the IR Czech test collection.
I closely work not only with my students, but also with other Czech and foreign teams, such as the University or West Bohemia in the Czech Republic, Center for Speech and Language Processing at the JHU, the CLEAR lab at CU Boulder, Linguistic Data Consortium, the European Language Resources Association (ELRA/ELDA), and several European Universities on EU projects