The LT³ team is a young but highly active research group at Ghent University, Faculty of Arts, Humanities and Law. The actual team was not created until 2006 but nonetheless its members gained broad experience over the years in the field of NLP research. The main focus lies on corpus-based language and translation technology research. The team's research is built on the following three pillars:
- Terminology: terminological research and the construction of term banks. Given the observation that terminology management is crucial for translation and intelligent search applications, the LT³ team is involved in terminology extraction of domain-specific term banks. CvT-GenTerm, the faculty’s Terminology Centre, is affiliated to LT³ and plays an important role in manual construction of term banks.
- Corpora: the construction of monolingual and parallel corpora for linguistic and NLP research. Next to application development, corpora can also serve for different research purposes and as didactic material in education. LT³ currently participates in different corpus construction projects, which will be used to train inductive NLP systems for machine translation, named entity recognition, coreference resolution, etc.
- Language technology: language and translation technology research and the development of corpus-based language and translation technology applications. The LT³ team conducts fundamental and applied research on the different aspects of Natural Language Processing (NLP) from a corpus-based, inductive perspective. The main focus lies on text mining, the automatic information extraction from unstructured text. Current research activities comprise both research in underlying NLP components such as word sense disambiguation, anaphora resolution, named entity recognition, etc. and end-user applications such as machine translation, terminology extraction, sentiment mining, automatic readability scoring, etc.