ClovaCall: Korean Goal-Oriented Dialog Speech Corpus for Automatic Speech Recognition of Contact Centers

Despite the advancement of ASR, however, most publicly trained from these speech data generally show poor recognition available call-based speech corpora such as Switchboard performance when applied to domain-specific tasks due to the are old-fashioned. Also, most existing call corpora are in English differences in their data distribution and vocabularies. In particular, and mainly focus on open domain dialog or general scenarios AICC requires an accurate ASR model to ensure the precise such as audiobooks. Here we introduce a new large-scale intent classification or slot extraction [9] from user natural Korean call-based speech corpus under a goal-oriented dialog language utterances.