Context-aware Neural-based Dialog Act Classification on Automatically Generated Transcriptions
This work addresses dialog act classification for automated transcription systems, but it is incremental as it builds on existing neural and sequence modeling techniques.
The paper tackles dialog act classification on automatically generated transcriptions by proposing a CNN-CRF approach for context modeling, showing consistent accuracy improvements on MRDA and SwDA datasets and finding that End-to-End ASR systems are more suitable despite comparable word error rates.
This paper presents our latest investigations on dialog act (DA) classification on automatically generated transcriptions. We propose a novel approach that combines convolutional neural networks (CNNs) and conditional random fields (CRFs) for context modeling in DA classification. We explore the impact of transcriptions generated from different automatic speech recognition systems such as hybrid TDNN/HMM and End-to-End systems on the final performance. Experimental results on two benchmark datasets (MRDA and SwDA) show that the combination CNN and CRF improves consistently the accuracy. Furthermore, they show that although the word error rates are comparable, End-to-End ASR system seems to be more suitable for DA classification.