Reasoning on Relational Database and Its Respective Knowledge Graph: A Comparison of the Results

The use of relational databases has been a longstanding practice for data storage and information retrieval. Nevertheless, the emergence of knowledge graphs has led to a gradual displacement of relational databases in various domains. Thus, in this study we tend to analyze knowledge graphs’ reasoning abilities and perform a comparison with relational databases. To achieve optimal reasoning in data, we have incorporated neural networks algorithms. We tested aforementioned algorithms on a pre-existing knowledge graph along with its associated dataset and this paper presents the outcomes of our experiments. A subset of the HETIONET dataset related to cancer diseases, represented as a knowledge graph in neo4j, was carefully selected as the standard for our analysis to ensure objectivity and precision. The performance of the top algorithms employed in relational databases and knowledge graphs was our main goal. Our research contributes valuable insights to the comparative analysis of reasoning algorithms between relational databases and knowledge graphs. By shedding light on the strengths and weaknesses of different approaches, this study serves as a foundation for further advancements in database management and data extraction techniques. This study's findings can be used to create more complex algorithms and improve the functionality of knowledge graph-based systems.