Knowledge-Rich Morphological Priors for Bayesian Language Models
We present a morphology-aware nonparametric Bayesian model of language whose prior distribution uses manually constructed finitestate transducers to capture the word formation processes of particular languages. This relaxes the word independence assumption and enables sharing of statistical strength across, for example, stems or inflectional paradigms in different contexts. Our model can be used in virtually any scenario where multinomial distributions over words would be used. We obtain state-of-the-art results in language modeling, word alignment, and unsupervised morphological disambiguation for a variety of morphologically rich languages.