Hyungsuk Ji's Home Page


Brief Introduction of Myself

My main research interest is in modelling natural language. As demonstrated in recent studies, an understanding of the language-related cognitive capacities of humans that goes beyond the scope of natural language is crucial for designing a reliable, high-performance natural-language model.

I have developed a computational model that is capable of automatically categorizing contextually-related words in a semantic representation, once it has been trained on a very large corpus. It could (1) serve as a model to account for human concepts and categorization processes, and (2) be used as a machine translation tool or other NLP tasks. The public version of the model can be tested on the Internet (the website address can be found in references).

Upon completion of my thesis on computational modelling for my master's degree in physical chemistry, I decided to attack more intelligent modelling and make it my life-long interest. I began my study of cognitive science in France (licence, maîtrise, D.E.A. and doctor's degrees.). As shown in my resumé, my studies have focused on language-related modelling, which I consider the most interesting and challenging subject in cognitive science.

I am planning to develop a semantic theory on the organization and evolution of words’ meaning that could be supported by a computational model that learns this capability automatically.

Hyungsuk Ji

LED (Language and Dialogue Team)
LORIA, Campus Scientifique - BP 239
F-54506 Vandoeuvre Cedex

tel.: +33 (0)3 8359 2047
fax: +33 (0)3 8341 3079
e-mail: Hyung-suk.JI__at__loria.fr



Institut National Polytechnique de Grenoble

Cognitive Science. Dissertation: A Computational model for word sense representation using contextual relations




Institut National Polytechnique de Grenoble

Cognitive Science. Thesis: A Computational model for the representation of lexicons in phonological-semantic space




University of Bordeaux II

Cognitive Science. Thesis: Natural language production: simulation with IAC neural network (Interactive Activation and Competition network)




University of Bordeaux II

Cognitive Science




Seoul National University (the most prestigious institution of higher education in South Korea)

Chemistry. Thesis: Computational molecular dynamics simulation

South Korea


Academic position

Postdoctoral fellow at INRIA, France (2004 – present)

Fellowship awarded by INRIA (National Institute for Research in Computer Science and Automatic Control Engineering)

Collaborative experience

Research at LATL (Natural Language Analysis and Technology Lab) (2001 – 2003, discontinuous), Department of Linguistics, University of Geneva, Switzerland

Eurodoc Scholarship awarded


Hyungsuk Ji, Sabine Ploux and Eric Wehrli. (2003). Lexical knowledge representation with contexonyms. In Proceedings of the 9th MT summit, 194-201. (PDF, ps)

Hyungsuk Ji and Sabine Ploux. (2003). Automatic contexonym organizing model (ACOM). In Proceedings of the 25th Annual Conference of the Cognitive Science Society, 622-627. (PDF, ps)

Hyungsuk Ji and Sabine Ploux. (2003). A mental lexicon organization model.In Proceedings of the 4th Joint International Conference on Cognitive Science , 240-245. (PDF, ps)

Sabine Ploux and Hyungsuk Ji. (2003). A model for matching semantic maps between languages (French/English, English/French). Computational Linguistics , 29(2):155-178. (MIT Press(CL), PDF, ps)

Research interests

Technical skills / Academic service


Résumé (U.S.A.) / Curriculum vitæ (U.K.) in PDF format

In Korean