A Bayesian approach to language learning
Add to your list(s)
Download to your calendar using vCal
If you have a question about this talk, please contact Zoubin Ghahramani.
Children learning language are faced with a difficult task: to identify the correct generalizations to draw from highly structured linguistic data, with little explicit feedback. What constrains the learner to generalize appropriately? Research into this question is useful both for understanding human cognition, and for improving unsupervised language learning in machines. In this talk, I discuss the Bayesian approach to language acquisition and describe a nonparametric Bayesian modeling framework that can be used to examine a variety of different language learning tasks. I provide examples from word segmentation (identifying words from continuous text or speech) and morphology (identifying stems and suffixes), showing that these models are successful both in learning from corpora and at modeling human experimental data.
This talk is part of the Machine Learning @ CUED series.
This talk is included in these lists:
Note that ex-directory lists are not shown.
|