Multiple Generative Models Ensemble for Knowledge-Driven Proactive Human-Computer Dialogue Agent
Published in
arXiv.org(2019)External Links:
TL
TL;DR
An end-to-end multi-turns proactive dialogue generation agent was established with the aid of data augmentation techniques and variant encoder-decoder structure designs and a rank-based ensemble approach was developed for boosting performance.
Abstract
Multiple sequence to sequence models were used to establish an end-to-end multi-turns proactive dialogue generation agent, with the aid of data augmentation techniques and variant encoder-decoder structure designs. A rank-based ensemble approach was developed for boosting performance. Results indicate that our single model, in average, makes an obvious improvement in the terms of F1-score and BLEU over the baseline by 18.67% on the DuConv dataset. In particular, the ensemble methods further significantly outperform the baseline by 35.85%.
References16 items
1
Proactive Human-Machine Conversation with Explicit Conversation Goal
2
EDA: Easy Data Augmentation Techniques for Boosting Performance on Text Classification Tasks
3
An Ensemble of Retrieval-Based and Generation-Based Human-Computer Conversation Systems
4
Directional Skip-Gram: Explicitly Distinguishing Left and Right Context for Word Embeddings
5