Bayesian Network based Cost Benefit Factor Inference in E-Services

AbsIractThis paper applies Bayesian network tecbnique to model and inference tbe uncertain relationsbips among cost factors and benefit factors in E-services. A cost-benefit factor-relation modelproposed inour previous study isconsidered as domain knowledgeand tbe data collectedtbrougb a survey is as evidence to conduct inference. Tbrough calculating conditional probability distribution among factors and conducting inference, this paper identifies tbat certain cost factors are significantly more important tban others to certain benefit factors. In particular, this study found tbat 'increased investment in maintaining E-services' would significantly contribute to 'enbancing perceived company image' and 'gaining competitive advantages', and 'increased investment in staff training' would significant contribute to 'realizing business strategies'. These results bave the potential to improve the strategic planning of companies by determining more effective investment areas and adopting more suitable development activities wbere &services are concerned.