内容摘要：Data is becoming an increasingly important asset in today’s organizations. Various challenges in the age of big data, such as high volume and high velocity, call for efficient database maintenance policies. This research focuses on deriving an optimal update policy for consolidated database systems. By trading off update cost against data staleness cost, we propose a time-based dynamic update (TDU) policy, which evaluates the system state at predetermined checkpoints and updates the database only if given staleness thresholds are crossed. An optimal TDU policy is developed for both finite and infinite time horizons. Although several database update and knowledge refreshment policies have been proposed in the prior literature, the TDU policy retains their advantages and overcomes their inflexibility in that assessment and update are easy to schedule and update operation is run only if necessary. Experimental results show that the TDU policy consistently outperforms the benchmark periodic policy, leading to substantial cost savings. The performance gap is greater when data errors arrive less frequently but carry higher unit staleness cost, or when queries arrive more frequently and suffer a higher loss from data errors.

报告人简介：Zhengrui Jiang is the Thome Professor in Business and associate professor of information systems at the College of Business, Iowa State University. He received his PhD in Management Science/Information Systems from the University of Texas at Dallas. His primary research interests include business intelligence/analytics, diffusion of innovations, and economics of information goods. He has published in leading academic journals includingManagement Science, MIS Quarterly, Information Systems Research, IEEE Transactions on Knowledge and Data Engineering, INRORMS Journal on Computing,andJournal of Management Information Systems. His research has been funded by U.S. Agency for International Development, and National Natural Science Foundation of China. He currently serves as an associate editor for MIS Quarterly. He previously served as guest editors for Production and Operations Management and IEEE Transactions on Management Information Systems. He also served as a program co-chair for the 2014 Midwest Association of Information Systems Conference and the 2015 Big XII+ MIS Research Symposium. He will be co-chairing the 2018 Workshop on Information Technology and Systems.