Loading...
Please wait, while we are loading the content...
Similar Documents
Evaluation Metrics For Language Models (1998)
| Content Provider | CiteSeerX |
|---|---|
| Author | Chen, Stanley Rosenfeld, Ronald Beeferman, Douglas |
| Description | In DARPA Broadcast News Transcription and Understanding Workshop |
| Abstract | The most widely-used evaluation metric for language models for speech recognition is the perplexity of test data. While perplexities can be calculated efficiently and without access to a speech recognizer, they often do not correlate well with speech recognition word-error rates. In this research, we attempt to find a measure that like perplexity is easily calculated but which better predicts speech recognition performance. We investigate two approaches; first, we attempt to extend perplexity by using similar measures that utilize information about language models that perplexity ignores. Second, we attempt to imitate the word-error calculation without using a speech recognizer by artificially generating speechrecognition lattices. To test our new metrics, we have built over thirty varied language models. We find that perplexity correlates with word-error rate remarkably well when only considering n-gram models trained on in-domain data. When considering other types of models, our nove... |
| File Format | |
| Publisher Date | 1998-01-01 |
| Access Restriction | Open |
| Subject Keyword | N-gram Model Speechrecognition Lattice Speech Recognition Utilize Information Predicts Speech Recognition Performance Speech Recognition Word-error Rate Thirty Varied Language Model Word-error Calculation Language Model New Metric Word-error Rate In-domain Data Speech Recognizer Test Data Perplexity Ignores Widely-used Evaluation Similar Measure Evaluation Metric |
| Content Type | Text |
| Resource Type | Conference Proceedings |
