Well Begun is Half Done: Generator-agnostic Knowledge Pre-Selection for Knowledge-Grounded Dialogue
This addresses the challenge of reducing the burden on large language models in dialogue systems, though it is incremental as it builds on existing knowledge selection approaches.
The paper tackles the problem of accurate knowledge selection for knowledge-grounded dialogue systems by proposing GATE, a generator-agnostic method that selects context-related knowledge before generation, which improves response informativeness in experiments with models like ChatGPT.
Accurate knowledge selection is critical in knowledge-grounded dialogue systems. Towards a closer look at it, we offer a novel perspective to organize existing literature, i.e., knowledge selection coupled with, after, and before generation. We focus on the third under-explored category of study, which can not only select knowledge accurately in advance, but has the advantage to reduce the learning, adjustment, and interpretation burden of subsequent response generation models, especially LLMs. We propose GATE, a generator-agnostic knowledge selection method, to prepare knowledge for subsequent response generation models by selecting context-related knowledge among different knowledge structures and variable knowledge requirements. Experimental results demonstrate the superiority of GATE, and indicate that knowledge selection before generation is a lightweight yet effective way to facilitate LLMs (e.g., ChatGPT) to generate more informative responses.