University of Geneva
40 bd. du Pont-d'Arve
CH-1211 Geneva 4
Tel: +41 22 379 8691
Fax: +41 22 379 8689
Office number: 6337
Background and Research Interests
I have been working as a teaching assistant and researcher at ISSCO/TIM, since October 2004. I hold an MSc in Translation studies (with minors in economics, international shipping, journalism and art history) from Turku University, Finland, as well as an MSc in Multilingual Natural Language Processing from TIM, at the ETI, Geneva University. I'm currently pursuing Phd at TIM.
I'm interested in how the development of rule-based applications could be made less laborious, especially in a multilingual context. Rule-based systems like speech recognisers and MT systems are based on manually developed language resources, including grammars, lexica etc. My PhD investigates the option of a shared grammar that can be used for different types of languages (Finnish, English, Japanese) and for different tasks (speech recognition, analysis and generation) inside the multilingual spoken language translator MedSLT.
Brief descriptions of some of the projects that I've worked on:
- MedSLT - the MedSLT system is a multilingual speech- to-speech translation system for the medical domain. I have been working on developing language ressources for Finnish, and I've developed a shared English, Japanese, Finnish and Greek MedSLT grammar for purposes of my Phd.
- Regulus - Regulus is an open-source toolkit for building rule-based spoken dialogue systems.
- TransTec - the TransTec project was developing e-learning tools and course material for translation courses.
I am also helping organise the Workshop on Grammar Engineering Across Frameworks - GEAF 2009, which will be held at the ACL/IJCNLP 2009 Conference in Singapore.
I am involved in the following courses:
- Gestion et recherche documentaire (BA)
- Nouvelles technologies pour la traduction (MA)
- Informatique et communication (BA) (2004-2007)
A full list of publications can be found here.
- Multilingual Grammar Resources in Multilingual Application Development, Santaholma M., in: Proceedings of Workshop on Grammar Engineering Across Frameworks, GEAF. Manchester, UK.
- A knowledge-modeling approach for multilingual Regulus lexica , Santaholma M. and Chatzichrisafis N., n: proceedings of LREC 2008, Marrakech. Morocco.
- Making Speech Look Like Text in the Regulus Development Environment, Kron E., Rayner M., Santaholma M., Bouillon P. and Lisowska A., in: Proceedings of Workshop on Grammar Engineering Across Frameworks (GEAF2008), Coling 2008, Manchester, England.
- Grammar sharing techniques for rule-based multilingual NLP systems, Santaholma M., in: Proceedings of NODALIDA 2007, the 16th Nordic Conference of Computational Linguistics, May 25-26 2007, Tartu, Estonia.
- Regression Testing Grammar-Based Systems, Chatzichrisafis N., Crouch D., King T. H. , Nairn R. ,Rayner M. and Santaholma M., In: Tracy Holloway King and Emily Bender (Eds.): Proceedings of Workshop Grammar Engineering across Frameworks 2007. CSLI Publications, Stanford, California, pp. 128-143, 2007.
(Full list of publications related to MedSLT and Regulus projects can be found on the project pages.)