Abstract
In this work, a knowledge-based system for the benchmarking of public university libraries was developed in order to provide more useful information for decision-making and process benchmarking. This knowledge-based system converts numerical data into information that can be used to evaluate relative efficiency of public university libraries in Taiwan. We propose an integrated framework for the benchmarking wheel and knowledge-based system, including a database management subsystem, model base subsystem, knowledge acquisition subsystem, and dialogue subsystem. In the model-base, we use the DEA models, including the CCR and BCC models, in order to shed new light on the operational efficiency of public libraries in Taiwan. Finally, we provide strategies for improving inefficient public university libraries