The inability of large linguistic models to attribute their statements to external knowledge and their tendency to hallucinate makes it difficult to trust their answers. Even humans are prone to making factual errors in their writing. Therefore, verifying the accuracy of textual information, whether generated by large language models or curated by humans, is an important task. However, manually validating and correcting factual errors tends to be a tedious and laborious process. In this paper, we propose FLEEK for automatic fact checking and correction. FLEEK automatically extracts factual climates within the text, retrieves relevant evidence for each claim from various external knowledge sources, and then evaluates the factual status of each claim based on the recovered evidence. The system also automatically corrects factual errors detected in claims based on the evidence recovered. Experiments show that FLEEK is able to exhaustively extract factual claims, correctly determine their factual status, and propose meaningful corrections based on the recovered evidence.
