Knowledge Graph Based Synthetic Corpus Generation for Knowledge-Enhanced Language Model Pre-training

TL;DR

This paper proposes converting large-scale knowledge graphs into natural language text to enhance language model pre-training, improving factual accuracy and reducing toxicity in knowledge-intensive tasks.

Contribution

It introduces a novel method to verbalize the entire Wikidata knowledge graph, enabling seamless integration into language models for better performance.

Findings

Significant improvements in open domain QA accuracy

Enhanced performance on LAMA knowledge probe

Reduced toxicity in language model outputs

Abstract

Prior work on Data-To-Text Generation, the task of converting knowledge graph (KG) triples into natural text, focused on domain-specific benchmark datasets. In this paper, however, we verbalize the entire English Wikidata KG, and discuss the unique challenges associated with a broad, open-domain, large-scale verbalization. We further show that verbalizing a comprehensive, encyclopedic KG like Wikidata can be used to integrate structured KGs and natural language corpora. In contrast to the many architectures that have been developed to integrate these two sources, our approach converts the KG into natural text, allowing it to be seamlessly integrated into existing language models. It carries the further advantages of improved factual accuracy and reduced toxicity in the resulting language model. We evaluate this approach by augmenting the retrieval corpus in a retrieval language model and showing significant improvements on the knowledge intensive tasks of open domain QA and the LAMA knowledge probe.