摘要

In recent years, many companies have adopted outsourcing for their IS system maintenance. Outsourcing can be expensive, so many companies need to know the optimal effort of IT maintenance. Also, IT maintenance service providers try to come up with optimal man-months prediction and provide reliable service to clients, but still remain profitable. Problems with the recent trend of IT outsourcing are that clients become sensitive about the quality of IT outsourcing services and worry about the cost. Therefore, outsourcing service providers cannot be successful as service providers in the IS outsourcing market without convincing the client of the objectivity and transparency of IT outsourcing-related costs. This study investigates how to effectively determine proper man-months for the IT maintenance service to clients. To that end, this study derived various types of variables from one of the companies that outsources its IT maintenance service. To predict and verify the proper size of man-month, we employed Bayesian network (BN) classifier as a way to analyze methods. A selected BN structure-leaning algorithm and Markov blanket in this empirical experiment show that our method adapted well to man-month prediction study in the IS maintenance outsourcing domain. This paper also discusses some of limitations and future research in this field.

  • 出版日期2011-9