Deep Reinforcement Learning for Chatbots Using Clustered Actions and Human-Likeness Rewards
arXiv:1908.10331
Abstract
Training chatbots using the reinforcement learning paradigm is challenging due to high-dimensional states, infinite action spaces and the difficulty in specifying the reward function. We address such problems using clustered actions instead of infinite actions, and a simple but promising reward function based on human-likeness scores derived from human-human dialogue data. We train Deep Reinforcement Learning (DRL) agents using chitchat data in raw text---without any manual annotations. Experimental results using different splits of training data report the following. First, that our agents learn reasonable policies in the environments they get familiarised with, but their performance drops substantially when they are exposed to a test set of unseen dialogues. Second, that the choice of sentence embedding size between 100 and 300 dimensions is not significantly different on test data. Third, that our proposed human-likeness rewards are reasonable for training chatbots as long as they use lengthy dialogue histories of >=10 sentences.
arXiv commentIn International Joint Conference of Neural Networks (IJCNN), 2019
References in corpus (6)
- Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
- Distributed Representations of Words and Phrases and their Compositionality
- A Neural Conversational Model
- Why We Need New Evaluation Metrics for NLG
- Multiresolution Recurrent Neural Networks: An Application to Dialogue Response Generation
- A Deep Reinforcement Learning Chatbot (Short Version)