Efficient language model development for spoken dialogue recognition and its evaluation on operator’s speech at call centers

Efficient language model development for spoken dialogue recognition and its evaluation on operator’s speech at call centers Kiyokazu MIKI, Kaichiro HATAZAKI, and Hiroaki HATTORI Media and Information Research Laboratories, NEC Corporation 1753 Shimonumabe, Nakahara-Ku, Kawasaki, Kanagawa 211-8666, JAPAN k-miki@bq.jp.nec.com, k-hatazaki@ah.jp.nec.com, h-hattori@bp.jp.nec.com Abstract. While a language model for recognition of spoken dialogue is ideally built from a very large, specific-task-oriented corpus, a great amount of time and effort is required to develop such a corpus, and this involves both the audio recording and written transcription of large amounts of speech data. Training data for a language model should match the target task in both topic and style. What is needed, then, is a method to utilize previously existing spoken dialogue corpora that are not necessarily related to the specific target-task. Such corpora would be combined with documents related to the topic of the target-task to develop a language model for the target spoken-dialogue. In this paper, we propose a method for combining previously existing corpora with key phrases (i.e. phrases that contain keywords) extracted from task related documents. Even though the added data is from documents related to the target dialogue, since it consists of key phrases, stylistic differences (between document data and the actual dialogue to which the model will be applied) are not a problem. We have produced a model using this method and have evaluated it in use on actual spoken dialogue collected at call centers. Experimental results show that a relative 13% reduction in word error rate could be achieved with the addition of key phrases. This performance is nearly as good as that which would be achieved on the basis of a large, expensive transcript-corpus, and the cost of producing the key phrase data is essentially negligible. Such cost reduction achieved by our method will enable speech recognition applications to be more widely used. Keywords: Spoken dialogue recognition, Language model adaptation, Call center. 1 Introduction Call centers are particularly appropriate for use as speech recognition application domains because almost all of their operations are based on spoken dialogue. The many attempts made to date to apply speech recognition technology to call center applications have mostly been based on IVR (Interactive Voice Response) including automatic call-routing, which is designed for taking over human operator work in recognizing and responding to customer speech. On the other hand, we are developing another speech recognition application that recognizes operator speech and helps human operator work. For example, product names are extracted from speech recognition results of the operator speech and the information about them is displayed to the operator. An application focusing on operator speech would seem more promising because of the advantage it has for ease of speech recognition. While customer speech is telephone-transmitted, operator speech can be recorded directly from a headset microphone, and the sound quality is much higher. Further, the enunciation of professionally trained operators can be expected to be clearer from the beginning. Even for the recognition of operator speech at call centers, however, a …

Content truncated to 500 words