Dual-Feedback Knowledge Retrieval for Task-Oriented Dialogue Systems
This work addresses efficiency and scalability issues for developers of task-oriented dialogue systems, representing an incremental improvement by adapting open-domain QA techniques.
The paper tackles the scalability challenge of integrating knowledge retrieval and response generation in task-oriented dialogue systems by proposing a retriever-generator architecture with a dual-feedback mechanism for training the retriever using pseudo-labels from the generator, achieving superior performance on three benchmark datasets.
Efficient knowledge retrieval plays a pivotal role in ensuring the success of end-to-end task-oriented dialogue systems by facilitating the selection of relevant information necessary to fulfill user requests. However, current approaches generally integrate knowledge retrieval and response generation, which poses scalability challenges when dealing with extensive knowledge bases. Taking inspiration from open-domain question answering, we propose a retriever-generator architecture that harnesses a retriever to retrieve pertinent knowledge and a generator to generate system responses.~Due to the lack of retriever training labels, we propose relying on feedback from the generator as pseudo-labels to train the retriever. To achieve this, we introduce a dual-feedback mechanism that generates both positive and negative feedback based on the output of the generator. Our method demonstrates superior performance in task-oriented dialogue tasks, as evidenced by experimental results on three benchmark datasets.