Corpus ID: 115131886

Enhanced Sequential Representation Augmented with Utterance-level Attention for Response Selection

  title={Enhanced Sequential Representation Augmented with Utterance-level Attention for Response Selection},
  author={Taesun Whang and Dongyub Lee and Chanhee Lee and Heuiseok Lim},
Response selection is the task of choosing a correct response from a set of candidates. Predicting the next utterance is important in a dialog system since it helps the conversational agents converse consistently with humans. We propose an end-to-end response selection model for one of the tasks in the seventh Dialog System Technology Challenges (DSTC 7). The proposed model is based on word-level ESIM (Enhanced Sequential Inference Model) augmented with utterance-level attention. In the dialog… Expand

Figures and Tables from this paper


Multi-view Response Selection for Human-Computer Conversation
Enhance word representation for out-of-vocabulary on Ubuntu dialogue corpus
Training End-to-End Dialogue Systems with the Ubuntu Dialogue Corpus
Dialog System Technology Challenge 7
The Ubuntu Dialogue Corpus: A Large Dataset for Research in Unstructured Multi-Turn Dialogue Systems
An Information Retrieval Approach to Short Text Conversation
Deep contextualized word representations
Enhanced LSTM for Natural Language Inference
Attention is All you Need
Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation