Enhanced Graph Transforming Algorithm to solve transitive dependency between vertices

Abstract

Rapid and vast growth of data volume triggers a need for data management system with good scalability, availability and reliability. NoSQL database comes as a database management system developed to answer the need for such large scale and distributed data management. It is considered more effective and efficient compared to relational database. This database characteristics are very different than relational database. Therefore, a good method is necessary in order to transform data from relational database to NoSQL database. The Graph Transforming Algorithm is one of the simple methods and easy to use for that purpose. However, the result of conversion using this method remains with a considerably high level of data redundancy, making it necessary to further develop the algorithm. This research proposes a new algorithm called Enhanced Graph Transforming Algorithm, a development of Graph Transforming Algorithm by adding rules to eliminate transitive dependency between vertices prior to joining the vertices. Based on the test conducted in the research, the Enhanced Graph Transforming Algorithm is proven capable of reducing the number of redundant attributes and suppressing waste of space on data storage by up to 39.68% lower than the result of transformation generated using Graph Transforming Algorithm.