A Simple and Efficient Multi-Task Learning Approach for Conditioned Dialogue Generation
arXiv:2010.11140
Abstract
Conditioned dialogue generation suffers from the scarcity of labeled responses. In this work, we exploit labeled non-dialogue text data related to the condition, which are much easier to collect. We propose a multi-task learning approach to leverage both labeled dialogue and text data. The 3 tasks jointly optimize the same pre-trained Transformer -- conditioned dialogue generation task on the labeled dialogue data, conditioned language encoding task and conditioned language generation task on the labeled text data. Experimental results show that our approach outperforms the state-of-the-art models by leveraging the labeled texts, and it also obtains larger improvement in performance comparing to the previous methods to leverage text data.
arXiv commentAccepted as NAACL 2021 Long Paper (Oral)
References in corpus (16)
- ERNIE: Enhanced Representation through Knowledge Integration
- Wizard of Wikipedia: Knowledge-Powered Conversational agents
- Topic Aware Neural Response Generation
- TransferTransfo: A Transfer Learning Approach for Neural Network Based Conversational Agents
- Sentence Encoders on STILTs: Supplementary Training on Intermediate Labeled-data Tasks
- Learning Discourse-level Diversity for Neural Dialog Models using Conditional Variational Autoencoders
- ERNIE: Enhanced Language Representation with Informative Entities
- DialogWAE: Multimodal Response Generation with Conditional Wasserstein Auto-Encoder
- DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generation
- Generating Wikipedia by Summarizing Long Sequences
- Multi-Task Learning for Speaker-Role Adaptation in Neural Conversation Models
- PLATO-2: Towards Building an Open-Domain Chatbot via Curriculum Learning
- CAiRE: An Empathetic Neural Chatbot
- A Pre-training Based Personalized Dialogue Generation Model with Persona-sparse Data
- Generating Multiple Diverse Responses with Multi-Mapping and Posterior Mapping Selection
- Open-Domain Dialogue Generation Based on Pre-trained Language Models