Textual Entailment for Effective Triple Validation in Object Prediction
This work addresses the problem of unreliable fact extraction in knowledge base population for AI researchers and practitioners, offering an incremental improvement over existing methods.
The paper tackled the brittleness and hallucination issues in prompt-based fact retrieval from language models by using textual entailment for triple validation, resulting in improved predictions across different training regimes and effective validation of facts from other sources.
Knowledge base population seeks to expand knowledge graphs with facts that are typically extracted from a text corpus. Recently, language models pretrained on large corpora have been shown to contain factual knowledge that can be retrieved using cloze-style strategies. Such approach enables zero-shot recall of facts, showing competitive results in object prediction compared to supervised baselines. However, prompt-based fact retrieval can be brittle and heavily depend on the prompts and context used, which may produce results that are unintended or hallucinatory.We propose to use textual entailment to validate facts extracted from language models through cloze statements. Our results show that triple validation based on textual entailment improves language model predictions in different training regimes. Furthermore, we show that entailment-based triple validation is also effective to validate candidate facts extracted from other sources including existing knowledge graphs and text passages where named entities are recognized.