A hybrid deep learning and ontology-driven approach to perform business process capability assessment

Abstract

Enterprises are constantly transforming to adapt to an ever-changing and competitive environment. In this context, assessments allow to understand the state of different organisational aspects before performing transformation activities. One of these aspects is the capability of business processes. Evaluating the quality of business processes is relevant to guide improvement initiatives, considering that the way that processes are designed and executed in organisations has direct impact on the quality of products and services. However, assessments are expensive in terms of resources if they are performed by humans. In this sense, recent trends in Artificial Intelligence provide means to improve process capability assessment through the automation of some of its tasks. Following this line, this work presents a method to perform process capability assessment using raw text as input data with the aid of a smart system, able to reduce the need of human intervention to provide reliable assessment results. For this purpose, we introduce a hybrid approach to perform assessments in enterprises using text data as assessment evidence. The method combines the Long Short-Term Memory Network (LSTM) approach and the use of an Ontology named Process Capability Assessment Ontology (PCAO), which also contains a set of rules to calculate process attribute ratings, capability levels, among other aspects. The approach is grounded on the Smart Assessment Framework, a conceptual model devised to guide the development of intelligent assessments in enterprises. We introduce a demonstration of the assessment of a process based on the management of chemical samples from a research institute.