Abstract | ||
---|---|---|
We address the problem of learning the mapping between words and their possible pronunciations in terms of sub-word units. Most previous approaches have involved generative modeling of the distribution of pronunciations, usually trained to maximize likelihood. We propose a discriminative, feature-rich approach using large-margin learning. This approach allows us to optimize an objective closely related to a discriminative task, to incorporate a large number of complex features, and still do inference efficiently. We test the approach on the task of lexical access; that is, the prediction of a word given a phonetic transcription. In experiments on a subset of the Switchboard conversational speech corpus, our models thus far improve classification error rates from a previously published result of 29.1% to about 15%. We find that large-margin approaches outperform conditional random field learning, and that the Passive-Aggressive algorithm for large-margin learning is faster to converge than the Pegasos algorithm. |
Year | Venue | Keywords |
---|---|---|
2012 | ACL | conditional random field learning,feature-rich approach,large-margin learning,classification error rate,previous approach,discriminative pronunciation modeling,large-margin approach,switchboard conversational speech corpus,passive-aggressive algorithm,pegasos algorithm,discriminative task |
Field | DocType | Volume |
Pronunciation,Speech corpus,Phonetic transcription,Computer science,Generative modeling,Natural language processing,Artificial intelligence,Discriminative model,Conditional random field,Lexical access,Inference,Speech recognition,Machine learning | Conference | P12-1 |
Citations | PageRank | References |
5 | 0.47 | 27 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Hao Tang | 1 | 5 | 0.47 |
Joseph Keshet | 2 | 925 | 69.84 |
Karen Livescu | 3 | 1254 | 71.43 |