Machine learning approach for model discovery and process enhancement using process mining techniques

Abstract

The prologue of computer system and networks for the business applications has started few decades ago and in a short span of time the entire business environment has been transformed in to Digital business environment and it has evolved to a spectacular dimension Start from Business automation Resource Planning Business intelligence From two thousand it is the era for business optimization Adaptation of digital environment for business processing is inevitable in the present scenario These digital business environment has brought immense convenient at various levels and equally triggered many serious challenges need to be addressed The top priority among the challenges is handling of Big data The transactional data stored in the data repositories for every business activity has grown over time and became unimaginable huge This thesis has given proper approach and technique to handle the situation cased by the digital business environment from the perspective of business process optimization The first chapter presents the need and necessity for the research and narrates the current situation prevailing in the modern digital business environment It has given options in addressing the challenges and justifies that providing only storage solution is not enough instead converting the challenge in to business understanding and improve the operational efficiency is wise It reveals the related and predecessor research contribution in the field of machine learning motor insurance and process mining It also explains that inductive method is most appropriate for the research situation The second chapter briefs about machine learning and its classifications based on the underlying learning strategies used in the representation of knowledge or skill acquired by the learner and the application domain in the performance system for which knowledge is acquired It also explains about Supervised and unsupervised machine learning types The third chapter explains process models particularly about Descriptive Prescriptive and Explanatory process models and its uses It also explains Discovery Conformance and Enhancement types of process mining The lifecycle model of Lasagna and spaghetti process and their respective opportunities in process mining are discussed The fourth chapter discusses the practical approach for the research work Insurance data particularly motor insurance data is taken as the data source for the research This chapter narrates the basic principles of insurance and its regulatory body IRDA Various types of motor vehicle and various types of motor insurances are discussed It ends by comparing the premium contribution of motor insurance with other general insurance in India newlineThe fifth chapter explains about model discovery and various standard model notations available It also explains the procedure followed while generating the personalized model starts from event log trace table and the model generator The Sixth chapter deals with the process mining tools particularly the open source tool ProM frame work and the ProM architecture The Meta Model of XES has been discussed Various Plug ins with description are listed The seventh chapter identifies the areas which need to be addressed and has scope of improvement through the feedback given by the customers Process conformance has been tested and enhanced using Woflan DISCO tool has been used for process analysis The eighth chapter briefs about the contribution of this thesis particularly about model generator various standard notations and process enhancement with regard to insurance claim process The final chapter touches upon the results of the research work and identifies the possibility for further research newline