RHO ($ρ$): Reducing Hallucination in Open-domain Dialogues with Knowledge Grounding
This addresses the issue of unfaithful responses in dialogue systems, which is crucial for reliable AI applications, though it is an incremental improvement over existing knowledge grounding methods.
The paper tackled the problem of hallucination in open-domain dialogue systems by proposing RHO, a method that uses knowledge graph embeddings and multi-hop reasoning to ground responses, resulting in a 17.54% reduction in hallucination on the FeQA metric.
Dialogue systems can leverage large pre-trained language models and knowledge to generate fluent and informative responses. However, these models are still prone to produce hallucinated responses not supported by the input source, which greatly hinders their application. The heterogeneity between external knowledge and dialogue context challenges representation learning and source integration, and further contributes to unfaithfulness. To handle this challenge and generate more faithful responses, this paper presents RHO ($ρ$) utilizing the representations of linked entities and relation predicates from a knowledge graph (KG). We propose (1) local knowledge grounding to combine textual embeddings with the corresponding KG embeddings; and (2) global knowledge grounding to equip RHO with multi-hop reasoning abilities via the attention mechanism. In addition, we devise a response re-ranking technique based on walks over KG sub-graphs for better conversational reasoning. Experimental results on OpenDialKG show that our approach significantly outperforms state-of-the-art methods on both automatic and human evaluation by a large margin, especially in hallucination reduction (17.54% in FeQA).