Vishleshan : performance comparison and programming process mining algorithms in graph-oriented and relational database query languages

Abstract

Information Systems today record the execution of activities into event logs. Process Mining is an area of research that deals with the study and analysis of various business processes based on these event logs. These event logs also record the performers of each of the activities. Mining social network using this information, understanding work-flow management and deriving relationships between actors based on different metrics viz. handover of task, subcontract, etc. is what constitutes Organizational Mining. Metrics based on Joint Activities and Metrics based on (possible) causality, commonly referred to as Similar-Task Algorithm and Subcontract Algorithm forms the basis of this paper. We present Cypher Query Language(Neo4j) and SQL (Structured Query Language) implementations of Similar-Task and Subcontract Algorithms. Graph Databases have shown to perform well in cases where information follows linked structure and needs to query to depth(s) of a hierarchical setup. We conduct an empirical study on a large real world data set to compare the performance of Neo4j against MySQL. We benchmark performance factors like query execution time, CPU usage and disk/memory space usage when implementing Similar-task and Subcontract algorithms in Cypher Query Language and SQL.