Context Matters in Semantically Controlled Language Generation for Task-oriented Dialogue Systems
This work addresses the need for more varied and contextually fitting responses in task-oriented dialogue systems, representing an incremental improvement over existing methods.
The paper tackles the problem of generating contextually appropriate responses in task-oriented dialogues by incorporating dialogue history and current utterance meaning, showing that using pre-trained models for context improves performance in both automatic metrics and human evaluation, with the immediate preceding user utterance being particularly important.
This work combines information about the dialogue history encoded by pre-trained model with a meaning representation of the current system utterance to realize contextual language generation in task-oriented dialogues. We utilize the pre-trained multi-context ConveRT model for context representation in a model trained from scratch; and leverage the immediate preceding user utterance for context generation in a model adapted from the pre-trained GPT-2. Both experiments with the MultiWOZ dataset show that contextual information encoded by pre-trained model improves the performance of response generation both in automatic metrics and human evaluation. Our presented contextual generator enables higher variety of generated responses that fit better to the ongoing dialogue. Analysing the context size shows that longer context does not automatically lead to better performance, but the immediate preceding user utterance plays an essential role for contextual generation. In addition, we also propose a re-ranker for the GPT-based generation model. The experiments show that the response selected by the re-ranker has a significant improvement on automatic metrics.