B659: Fall 2009
Multilingual Natural Language Processing
Indiana University School of Informatics and Computing
The multilingual world we live in and the multilingual Internet we now interact with present a range of challenges and opportunities, some of which are being addressed by the field of human language technology (natural language processing, computational linguistics). Information is often available in one language, but not another: this is a problem addressed by cross-linguistic information retrieval and machine(-assisted) translation. Multilingual digital resources and lexical databases such as WordNet permit the creation of multilingual ontologies, with numerous practical applications as well as implications for the nature of concepts. Finally, as in the past, there is great interest in the learning and teaching of second languages, processes that are increasingly aided by software with sophisticated linguistic models of the target and sometimes the source language.
This seminar will examine all of these areas of research, including both knowledge-based and statistical approaches.
Prerequisites: At least one graduate course in artificial intelligence or natural language processing (computational linguistics).
There is no textbook. Readings will come from recent journals and conference proceedings. There will be remedial lectures and lecture notes on some linguistic and computational linguistic topics. All readings will be available online.