Combining multiple information types in Bayesian word segmentation

Gabriel Doyle and Roger Levy

Humans identify word boundaries in continuous speech by combining multiple cues; existing state-of-the-art models, though, look at a single cue. We extend the generative model of Goldwater et al (2006) to segment using syllable stress as well as phonemic form. Our new model treats identification of word boundaries and prevalent stress patterns in the language as a joint inference task. We show that this model improves segmentation accuracy over purely segmental input representations, and recovers the dominant stress pattern of the data. Additionally, our model retains high performance even without single-word utterances. We also demonstrate a discrepancy in the performance of our model and human infants on an artificial-language task in which stress cues and transition-probability information are pitted against one another. We argue that this discrepancy indicates a bound on rationality in the mechanisms of human segmentation.