The BiFrameNet Project at the HKUST Human Language Technology Center
Bilingual resources are essential for Machine Translation and Cross-lingual Information Retrieval. However, current FrameNet is only for English, and other languages are either on a small scale or unfinished. We are going to construct Bilingal FrameNet, aiming at:
- Identify and quantify the mapping between semantic structures across language pairs for statistical Machine Translation systems
- Automatically create a multilingual FrameNet for semantic analysis
The BiFrameNet databse is composed of two parts. One is the bilingual ontology; the other is the set of bilingual example sentences.
- CHEN Benfeng and Pascale FUNG. "Automatic Construction of an English-Chinese Bilingual FrameNet". HLT/NAACL-2004: Proceedings of the Human Language Technology Conference of the North American Association for Computational Linguistics. Boston: May 2004.
The Hong Kong University of Science & Technology
HKUST, Clear Water Bay, Hong Kong
http://www.cs.ust.hk/~hltcLast updated: 2004.05.03 CHEN Benfeng