Cross-domain similarity assessment for workflow improvement to handle Big Data challenge in workflow management

Abstract

With the increasing of using workflow management systems workflow improvement becomes a new emerging problem. Many issues must be considered to handle all aspects of the workflow improvement. Workflows might become quite complex, especially when we move to Web3 (ubiquitous computing web). Workflows from different domains (e.g., scientific or business) have similarities and, more important, differences between themselves. Some concepts and solutions developed in one domain may be readily applicable to the other. In ubiquitous computing, multi-domain workflow data analysis might cause Big Data challenge. This paper investigates the problem of workflow improvement having an observed behavior (i.e., event logs). It proposes a cross-domain concept extraction by similarity assessment to solve some aspects of workflow improvement problem, and it has a new research effort at the intersection of workflow domains. Besides, the proposed technique is evaluated with the benefit of using Deep learning and Transfer learning. One of the greatest assets to use these both learning methods is analyzing a massive amount of data. Our results show that our proposed technique is effectively applicable for analyzing real-life huge data in workflow improvement.