Q-TOD: A Query-driven Task-oriented Dialogue System
This addresses the issue of domain adaptation and scalability in task-oriented dialogue systems for practical applications, representing a novel method rather than an incremental improvement.
The paper tackles the problem of task-oriented dialogue systems struggling with domain adaptation and large knowledge bases by introducing Q-TOD, a query-driven system that extracts natural language queries from dialogue context to retrieve knowledge, resulting in outperforming strong baselines and achieving state-of-the-art performance on three datasets.
Existing pipelined task-oriented dialogue systems usually have difficulties adapting to unseen domains, whereas end-to-end systems are plagued by large-scale knowledge bases in practice. In this paper, we introduce a novel query-driven task-oriented dialogue system, namely Q-TOD. The essential information from the dialogue context is extracted into a query, which is further employed to retrieve relevant knowledge records for response generation. Firstly, as the query is in the form of natural language and not confined to the schema of the knowledge base, the issue of domain adaption is alleviated remarkably in Q-TOD. Secondly, as the query enables the decoupling of knowledge retrieval from the generation, Q-TOD gets rid of the issue of knowledge base scalability. To evaluate the effectiveness of the proposed Q-TOD, we collect query annotations for three publicly available task-oriented dialogue datasets. Comprehensive experiments verify that Q-TOD outperforms strong baselines and establishes a new state-of-the-art performance on these datasets.