9th Annual Conference of the International Speech Communication Association

Brisbane, Australia
September 22-26, 2008

Comparing Word, Character, and Phoneme n-Grams for Subjective Utterance Recognition

Theresa Wilson (1), Stephan Raaijmakers (2)

(1) University of Edinburgh, UK; (2) TNO ICT, The Netherlands

In this paper, we compare the performance of classifiers trained using word n-grams, character n-grams, and phoneme n-grams for recognizing subjective utterances in multiparty conversation. We show that there is value in using very shallow linguistic representations, such as character n-grams, for recognizing subjective utterances, in particular, gains in the recall of subjective utterances.

Full Paper

Bibliographic reference.  Wilson, Theresa / Raaijmakers, Stephan (2008): "Comparing word, character, and phoneme n-grams for subjective utterance recognition", In INTERSPEECH-2008, 1614-1617.