Note: Your custom schedule will not be saved unless you create a new account or login to an existing account.
| Paper: | SP-L11.1 |
| Session: | Language Modeling and Search |
| Time: | Friday, May 21, 15:30 - 15:50 |
| Presentation: |
Lecture |
| Topic: |
Speech Processing: Language Modeling |
| Title: |
META-DATA CONDITIONAL LANGUAGE MODELING |
| Authors: |
Michiel Bacchiani; AT&T Labs - Research | | |
| | Brian Roark; AT&T Labs - Research | | |
| Abstract: |
Automatic Speech Recognition (ASR) often occurs in circumstances in which knowledge external to the speech signal, or meta-data, is given. For example, a company receiving a call from a customer might have access to a database record of that customer. Conditioning the ASR models directly on this information to improve the transcription accuracy is hampered because, generally, the meta-data takes on many values and a training corpus will have little data for each meta-data condition. This paper presents an algorithm to construct language models conditioned on such meta-data. It uses tree-based clustering of the the training data to automatically derive meta-data projections, useful as language model conditioning contexts. The algorithm was tested on a multiple domain voicemail transcription task. We compare the performance of an adapted system aware of the domain shift to a system that only has meta-data to infer that fact. The meta-data used were the callerID strings associated with the voicemail messages. The meta-data adapted system matched the performance of the system adapted using the domain knowledge explicitly. |
| |
| Back | |