Syllable weight encodes mostly the same information for English word segmentation as dictionary stress

Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 844–853, October 25-29, 2014, Doha, Qatar. c⃝2014 Association for Computational Linguistics Syllable weight encodes mostly the same information for English word segmentation as dictionary stress John K Pate Mark Johnson Centre for Language Technology Macquarie University Sydney, NSW, Australia {john.pate,mark.johnson}@mq.edu.au Abstract Stress is a useful cue for English word segmentation. A wide range of computa- tional models have found that stress cues enable a 2-10% improvement in segmen- tation accuracy, depending on the kind of model, by using input that has been anno- tated with stress using a pronouncing dic- tionary. However, stress is neither invari- ably produced nor unambiguously iden- tifiable in real speech. Heavy syllables, i.e. those with long vowels or syllable codas, attract stress in English. We de- vise Adaptor Grammar word segmentation models that exploit either stress, or sylla- ble weight, or both, and evaluate the util- ity of syllable weight as a cue to word boundaries. Our results suggest that sylla- ble weight encodes largely the same infor- mation for word segmentation in English that annotated dictionary stress does. 1 Introduction One of the first skills a child must develop in the course of language acquisition is the ability to seg- ment speech into words. Stress has long been recognized as a useful cue for English word seg- mentation, following the observation that words in English are predominantly stress-initial (Cutler and Carter, 1987), together with the result that 9- month-old English-learning infants prefer stress- initial stimuli (Jusczyk et al., 1993). A range of statistical (Doyle and Levy, 2013; Christiansen et al., 1998; B¨orschinger and Johnson, 2014) and rule-based (Yang, 2004; Lignos and Yang, 2010) models have used stress information to improve word segmentation. However, that work uses stress-marked input prepared by marking vowels that are listed as stressed in a pronouncing dic- tionary. This pre-processing step glosses over the fact that stress identification itself involves a non- trival learning problem, since stress has many pos- sible phonetic reflexes and no known invariants (Campbell and Beckman, 1997; Fry, 1955; Fry, 1958). One known strong correlate of stress in English is syllable weight: heavy syllables, which end in a consonant or have a long vowel, at- tract stress in English. We present experiments with Bayesian Adaptor Grammars (Johnson et al., 2007) that suggest syllable weight encodes largely the same information for word segmentation that dictionary stress information does. Specifically, we modify the Adaptor Grammar word segmentation model of B¨orschinger and Johnson (2014) to compare the utility of syllable weight and stress cues for finding word boundaries, both individually and in combination. We describe how a shortcoming of Adaptor Grammars prevents us from comparing stress and weight cues in combination with the full range of phonotactic cues for word segmentation, and design two experiments to work around this limitation. The first experiment uses grammars that provide parallel analyses for syllable weight and stress, and learns initial/non-initial phonotac- tic distinctions. In this first experiment, syllable …

Content truncated to 500 words