An Efficient, Probabilistically Sound Algorithm for Segmentation andWord Discovery
Machine Learning - Special issue on natural language learning
Unsupervised language acquisition
Unsupervised language acquisition
Segment predictability as a cue in word segmentation: application to modern Greek
SIGMorPhon '04 Proceedings of the 7th Meeting of the ACL Special Interest Group in Computational Phonology: Current Themes in Computational Phonology and Morphology
Hi-index | 0.00 |
Several computational simulations have been proposed for how children solve the word segmentation problem, but most have been tested only on a limited number of languages, often only English. In order to extend the cross-linguistic dimension of word segmentation research, a finite-state framework for testing various models of word segmentation is sketched, and a very simple cue is tested in this framework. Data is taken from Modern Greek, a language with phonological patterns distinct from English. A small-scale simulation shows using this cue performs significantly better than chance. The utility and flexibility of the finite-state approach is confirmed; suggestions for improvement are noted and directions for future work outlined.