愛荷華州立大學副教授Zhengrui Jiang：設計基于規則的智能專家系統應對說謊者

【主講】愛荷華州立大學副教授Zhengrui Jiang

【題目】設計基于規則的智能專家系統應對說謊者

【時間】2015年6月24日（周三）10.00-12.00

【地點】清華經管學院偉倫樓453

【語言】英文

【主辦】管理科學與工程系

【簡曆】Zhengrui Jiang老師的簡曆

Zhengrui Jiang is an associate professor of information systems and the inaugural holder of the endowed Thome Professorship in Business at the College of Business, Iowa State University. He received his PhD in Management Science from the University of Texas at Dallas. His research interests encompass business intelligence/analytics, diffusion of innovations, economics of IT, and software engineering and project management. His research has been funded by the U.S. Agency for International Development (USAID) and National Natural Science Foundation of China (NFSC). He has published in leading academic journals including Management Science, Information Systems Research, IEEE Transactions on Knowledge and Data Engineering, and Journal of Management Information Systems. He currently serves as an associate editor for MIS Quarterly and Information Technology and Management, and is a special issue senior editor for Production and Operations Management. He served as a program co-chair for the 2014 Midwest Association of Information Systems Conference and the 2015 Big XII+ MIS Research Symposium. In addition, he regularly serves on program committees of conferences/workshops such as WITS, CIST, and CSWIM.

Zhengrui Jiang, Associate Professor, Iowa State University: Designing Intelligent Rule-Based Expert Systems to Deal with Liars

【Speaker】Zhengrui Jiang, Associate Professor, Iowa State University

【Title】Designing Intelligent Rule-Based Expert Systems to Deal with Liars

【Time】Wednesday, June 24, 10.00-12.00

【Venue】Room 453, Weilun Building, Tsinghua SEM

【Language】English

【Organizer】Department of Management Science and Engineering

【Abstract】Input distortion is a common problem faced by expert systems, particularly those deployed with a Web interface. To address such a problem, we develop methods to distinguish liars from truth-tellers, and redesign rule-based expert systems to control the impact of input distortion by liars. The four methods developed in this study are termed Split Tree (ST), Consolidated Tree (CT), Value-based Split Tree (VST), and Value-based Consolidated Tree (VCT), respectively. Among them, ST and CT aim to increase a rule-based expert system’s accuracy of recommendations, and VST and VCT attempt to reduce the misclassification cost resulting from incorrect recommendations. We observe that ST and VST are less efficient than CT and VCT in that ST and VST require the verification of certain attribute values for all input scenarios, whereas the CT and VCT do not require value verification under certain scenarios. We conduct experiments to compare the performances of the four proposed methods and two existing methods, i.e., the traditional method that ignores input distortion and the knowledge modification (KM) method proposed in a prior research. The results show that CT and ST outperform the other methods in maximizing the recommendation accuracy, and VCT and VST outperform the other methods in minimizing the misclassification cost. Among the proposed methods, we find that CT consistently leads to a higher accuracy than ST, and VCT always results in a lower misclassification cost than VST. Therefore, CT and VCT should be the method of choice in dealing with users’lying behaviors. Furthermore, we find that the KM method, which considers input distortion but does not differentiate lairs from truth-tellers, is outperformed by not only the four proposed methods, but sometimes even by the method that ignores input distortions. This result further confirms the advantage of differentiating liars from truth-tellers when such distinctive groups exist in the population.