Post-hoc Utterance Refining Method by Entity Mining for Faithful Knowledge Grounded Conversations

TL;DR

This paper introduces REM, a post-hoc method that refines knowledge-grounded conversation responses by mining key entities to reduce hallucinations and improve faithfulness to source knowledge.

Contribution

The paper presents a novel post-hoc refinement approach that effectively reduces entity hallucinations in knowledge-grounded conversations.

Findings

REM decreases entity hallucination in generated responses.

The method improves faithfulness of responses to source knowledge.

Extensive experiments demonstrate REM's effectiveness and adaptability.

Abstract

Despite the striking advances in recent language generation performance, model-generated responses have suffered from the chronic problem of hallucinations that are either untrue or unfaithful to a given source. Especially in the task of knowledge grounded conversation, the models are required to generate informative responses, but hallucinated utterances lead to miscommunication. In particular, entity-level hallucination that causes critical misinformation and undesirable conversation is one of the major concerns. To address this issue, we propose a post-hoc refinement method called REM. It aims to enhance the quality and faithfulness of hallucinated utterances by refining them based on the source knowledge. If the generated utterance has a low source-faithfulness score with the given knowledge, REM mines the key entities in the knowledge and implicitly uses them for refining the utterances. We verify that our method reduces entity hallucination in the utterance. Also, we show the adaptability and efficacy of REM with extensive experiments and generative results. Our code is available at https://github.com/YOONNAJANG/REM.