Natural Language Processing

The Natural Language Processing group has been involved in human language technology research (primarily information extraction (IE)) since 1998, and by now, it has become one of the leading workshops of Hungarian computational linguistics. The Group is engaged in processing Hungarian and English texts. Its general objective is to develop language-independent or easily adaptable technologies.

IE aims at obtaining useful information, correlations automatically from running texts. It proves most expedient when lengthy texts are to be scanned in order to find the necessary information. In case of a suitably sized training database, the amount of human labor to be expended can be radically reduced by means of machine learning algorithms. Applications comprise the following important fields: information extraction from business news, biological publications, medical reports and from the Internet (forums, blogs, etc.). With the development of the manually annotated Szeged Corpus and TreeBank, as well as the Hungarian WordNet, SzegedNE and other corpora it has become possible to apply machine learning based methods for the syntactic and semantic analysis of Hungarian texts.

Senior and junior researchers, Ph.D. students, programmers and linguistic experts enrich the team with their knowledge. The group has more than 70 relevant international publications, it participated in 16 text processing related national RTD programs and delivered several industrial R&D solutions.

The Group has organised