Business rule optimisation: problem definition, proof-of-concept and application areas

Abstract

Business rules have been applied to a wide range of manufacturing and services organisations. Decisions around quality control, customer acceptance, and warranty claims are typical applications in day-to-day operation. They all have two things in common; there are multiple assessment criteria such as profit, revenue, and customer satisfaction, and the quality of the decisions made have an impact on the performance and sustainability of the organisation. This paper presents a solution to the novel problem of optimising the structure and parameters of automated business rules where there is the possibility to refer some decisions to a human expert. The difference here is that although the business rules are deterministic and repeatable, human decisions are generally neither. This research problem is multi-disciplinary, and the solution comprises elements of business process management, mathematical optimisation, simulation, machine learning, probability, and psychology. The paper describes a potential solution, some initial results when applied to a problem in the financial services sector and identifies further areas of application.