Multi-Granularity Representations of Dialog
arXiv:1908.09890
Abstract
Neural models of dialog rely on generalized latent representations of language. This paper introduces a novel training procedure which explicitly learns multiple representations of language at several levels of granularity. The multi-granularity training algorithm modifies the mechanism by which negative candidate responses are sampled in order to control the granularity of learned latent representations. Strong performance gains are observed on the next utterance retrieval task using both the MultiWOZ dataset and the Ubuntu dialog corpus. Analysis significantly demonstrates that multiple granularities of representation are being learned, and that multi-granularity training facilitates better transfer to downstream tasks.
Accepted as a long paper at EMNLP 2019
References in corpus (6)
- Sequence to Sequence Learning with Neural Networks
- Skip-Thought Vectors
- Machine Comprehension Using Match-LSTM and Answer Pointer
- TransferTransfo: A Transfer Learning Approach for Neural Network Based Conversational Agents
- A BERT Baseline for the Natural Questions
- Generative Encoder-Decoder Models for Task-Oriented Spoken Dialog Systems with Chatting Capability