The insurance industry must constantly adapt to digital trends and technologies as well as ever-changing customer expectations. For instance, claims processing should be designed in such a way that it is not only as efficient as possible, but also customer-oriented and customer-satisfiyng. In this thesis, we propose an optimization problem - the expert-based recommendation system (EBRS problem) - that can improve the claims processing by providing appropriate recommendations to clerks. In a first step, we estimate the (yet unknown) quality of a claims processing by taking into account the implicit knowledge of experts via utilizing adapted methods of Conjoint Analysis. In a second step, we use this expert knowledge to find suitable recommendations for the EBRS problem. This approach can be utilized to generate automated recommendations for actions to be taken by clerks to assist them in their everyday work tasks. The thesis addresses in particular the proof that the EBRS problem in general belongs to the complexity class of NP-complete problems. We finally investigate the EBRS problem in the context of complexity theory to identify cases for which the problem is solvable in polynomial time.
«The insurance industry must constantly adapt to digital trends and technologies as well as ever-changing customer expectations. For instance, claims processing should be designed in such a way that it is not only as efficient as possible, but also customer-oriented and customer-satisfiyng. In this thesis, we propose an optimization problem - the expert-based recommendation system (EBRS problem) - that can improve the claims processing by providing appropriate recommendations to clerks. In a firs...
»