Explicable knowledge graph (X-KG): generating knowledge graphs for explainable artificial intelligence and querying them by translating natural language queries to SPARQL

Knowledge graphs represent a potent instrument for the classification and exhibition of data, as they encompass a systematic approach for the containment and retrieval of multifarious datasets. In finance, the utilization of knowledge graphs for the organization of company-oriented data constitutes an invaluable source of insights, thus enabling informed decision-making. In a parallel, knowledge graph systems centered on COVID-19 within the healthcare sphere may assist medical professionals in the making of resolute choices. These applications highlight knowledge graphs’ ability to revolutionize decision-making procedures by providing a comprehensive comprehension of the given subject. To tackle this, we propose a solution that begets and implements knowledge graphs in two separate domains: finance and healthcare. To ensure the creation of explicable AI systems and improve the accessibility of information within these knowledge graphs, we introduce the conversion of natural language queries into SPARQL queries. By fine-tuning our model, we illustrate the system’s superior performance. Furthermore, we appraise the adequacy of the constructed knowledge graphs and contrast them with widely employed alternatives. Our work accentuates the adaptability of the proposed solution, as it can operate seamlessly with diverse datasets requiring minimal modifications.