Arabic Broadcast News Transcription Using a One Million Word Vocalized Vocabulary
| Autor: | Jean-Luc Gauvain, Lori Lamel, Abdelkhalek Messaoudi |
|---|---|
| Rok vydání: | 2006 |
| Předmět: |
Vocabulary
Arabic Computer science business.industry Speech recognition media_common.quotation_subject computer.software_genre Variety (linguistics) language.human_language Vowel language Language model Artificial intelligence Transcription (software) business computer Word (computer architecture) Natural language processing Natural language media_common |
| Zdroj: | ICASSP (1) |
| DOI: | 10.1109/icassp.2006.1660215 |
| Popis: | Recently it has been shown that modeling short vowels in Arabic can significantly improve performance even when producing a non-vocalized transcript. Since Arabic texts and audio transcripts are almost exclusively non-vocalized, the training methods have to overcome this missing data problem. For the acoustic models the procedure was bootstrapped with manually vocalized data and extended with semi-automatically vocalized data. In order to also capture the vowel information in the language model, a vocalized 4-gram language model trained on the audio transcripts was interpolated with the original 4-gram model trained on the (non-vocalized) written texts. Another challenge of the Arabic language is its large lexical variety. The out-of-vocabulary rate with a 65k word vocabulary is in the range of 4-8% (compared to under 1% for English). To address this problem a vocalized vocabulary containing over 1 million vocalized words, grouped into 200k word classes is used. This reduces the out-of-vocabulary rate to about 2%. The extended vocabulary and vocalized language model trained on the manually annotated data give a 1.2% absolute word error reduction on the DARPA RT04 development data. However, including the automatically vocalized transcripts in the language model reduces performance indicating that automatic vocalization needs to be improved.*†Visiting scientist from the Vecsys Company. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|