Clustering Public Library Users Based On Their Needs Using Artificial Neural Network, Analytical Hierarchy Process And Kano Model

Abstract Today, users have become a crucial factor to guidance of public libraries so public libraries require to know their users and plan for their satisfaction. This research proposes a hybrid model based on clustering method using neural networks, Analytical Hierarchy Process (AHP) and Kano model, try to segment public library users and obtain the value of each segment to provide appropriate services in order to satisfy them. The statistical population of this research consist of all public libraries users in the Fars province and a sample size of 180 is chosen by simple random sampling. Research tool is a questionnaire designed based on Kano model and the model parameters are determined according to the expertise of librarian. After data gathering, among the possible network architectures, self-organizing map (SOM) were chosen for segmenting or clustering users with the highest quality and using MATLAB software four clusters are identified for public library users in Fars province. Then clusters were prioritized using Analytic Hierarchy Process and finally by using the Kano model, needs of each cluster are identified and proper services for each cluster, are proposed. Using the results of this research public libraries can segment and identify customers and their satisfaction, the service would be more successful.