Domain-Independent True Fact Identification from Knowledge Graph

The trustworthiness of information in the Knowledge Graph (KG) is determined by the trustworthiness of information at the fact level. KGs are incomplete and noisy. Yet, most existing error detection approaches were applied to specific KGs. A large percentage of error detection approaches work well on DBpedia, particularly. However, we do not have a single KG containing all the information regarding the entity relations of a specific entity from any random class. The main objective of this research is to increase the trustworthiness of entity relations from KGs. In this paper, we propose a framework for identifying fact entity information that combines two independent approaches from knowledge graphs, ensuring the accuracy of triples. The first approach detects true facts of entity information from various KGs by integrating Linked Open Data (LOD), string similarity measures, and semantic similarity measures. Next, we propose an error detection and correction approach using RDF Reification on the integrated environment, independent of any particular KG. The research was conducted on related and diverse knowledge graphs, DBpedia, YAGO and Wikidata. In addition, the effectiveness of RDF reification for identifying true facts is evaluated on Wikidata on selected entities. The proposed framework provides a flexible framework for improving data quality across multiple KGs, enabling broader applicability in data integration and semantic search domains. Future work will explore extending this approach to deep learning models with additional features like entity type and path for error detection and correction in real-time KG applications.