Workshop on Data Science and Management 2023-09-18 为推动数据科学研究及其在商业、工程、社会管理等领域的应用，聚焦数据科学与管理科学交叉领域，突出运用数据科学理论方法，解决数字化、网络化、智能化发展中的管理难题和经济社会转型发展中的关键问题，由《数据科学与管理（英文）》编辑部主办，陕西省科学技术协会、西安交通大学管理学院、西安交通大学期刊中心、西安交通大学系统行为与管理教育部哲学社会科学实验室承办，中国系统工程学会青年工作委员会、中国优选法统筹法与经济数学研究会大数据与数据质量分会、the Association for Information Systems Special Interest Group on Information Systems in Asia Pacific协办的“数据科学与管理研讨会”将于2023年9月23日在西安交通大学（兴庆校区）管理学院召开，欢迎广大师生参会！ 会议日程 Keynote Speaker Zongben Xu Title: How to Simulate Learning Methodology？——On essence of Foundation Models Abstract: Zongben Xu is a professor in mathematics and computer science at Xi’an Jiaotong University. He received his Ph.D. degrees in mathematics from Xi’an Jiaotong University, China, in 1987. His current research interests include applied mathematics and mathematical methods of big data and artificial intelligence. He established the L(1/2) regularization theory for sparse information processing. He also found and verified Xu-Roach Theorem in machine learning, and established the visual cognition based data modelling principle, which have been widely applied in scientific and engineering fields. he initiated several mathematical theories, including the non-logarithmic transform based CT model, and ultrafast MRI imaging, which provide principles and technologies for the development of a new generation of intelligent medical imaging equipment. He is owner of the Hua Loo-keng Prize of Mathematics in 2022, Tan Kan Kee Science Award in Science Technology in 2018, the National Natural Science Award of China in 2007，and winner of CSIAM Su Buchin Applied Mathematics Prize in 2008. He delivered a 45-minute talk on the International Congress of Mathematicians 2010. He was elected as member of Chinese Academy of Science in 2011. Zongben Xu was the vice-president of Xi’an Jiaotong University. He currently makes several important services for government and professional societies, including the director for Pazhou Lab (Huangpu), director for the National Engineering Laboratory for Big Data Analytics, a member of National Big Data Expert Advisory Committee and the Strategic Advisory Committee member of National Open Innovation Platform for New Generation of Artificial Intelligence. Jae Kyu Lee Title: Preventive Cybersecurity as a Social Responsibility: The Bright Origin Approach Abstract: The cybercrimes and fakes on the Internet are so prolific that the cybersecurity has been the top concern to CIOs during last ten years consecutively according to the study of Society of Information Management (SIM). Since the complete protection of individual organization’s cybersecurity is nearly impossible to fulfill without the collective effort of mitigating the threat sources, we paid attention to the preventive cybersecurity paradigm and designed the Bright Internet Architecture with the Principle of Origin Responsibility. To validate the Bright Origin approach with the perspective of organization’s Social Responsibility, we identify the eight threat types and ten critical platforms that would create social damages beyond the organizational boundary. To measure the needs of preventive action on these risk sources, we survey the opinions of CIOs and IT experts who would understand the social responsibility aspect of cybersecurity for the future. According to this study, we found that they regard the preventive actions with social responsibility will mitigate the risk sources and will eventually contribute to building the safer environment for individual organizations. To motivate the efforts to realize the preventive actions, the direct benefits can be obtained by endowing the Certificate of Bright Origins that will create reputation from the customers and social community. Jae Kyu Lee is a Distinguished Professor of School of Management at Xi’an Jiaotong University and Professor Emeritus of Korea Advanced Institute of Science and Technology (KAIST). He has been professor of KAIST since 1985, and finished his tenure as HHI Chair Professor. He received fellow and LEO Award and served the President (2015-6) of Association for Information Systems. He is the founder of Principles for the Bright Internet and founded Bright Internet Research Center at KAIST and Xi’an Jiaotong University. He also founded the International Conference on Electronic Commerce (http://www.icec.net), and Bright Internet Global Symposium (http://www.brightinternet.org.). He received his Ph.D. in Information and Operations Management from the Wharton School, University of Pennsylvania in 1985. His research area covers AI, eCommerce, information systems, and Bright Internet, and published many academic papers in the international journals including MIS Quarterly, Information Systems Research, Management Science, Decision Support Systems, and Expert Systems with Applications, He served the founding editor-in-chief of Electronic Commerce Research Applications (Elsevier) and the editorial boards of many international academic journals. He currently serves as editorial advisor of ‘Data Science and Management’ and ‘Decision Support Systems’. Haibing Lu Title: Understanding and Addressing Bias and Privacy Challenges in Large Language Models Abstract: Large language models have reshaped the AI landscape, yet they cast shadows of bias and privacy concerns that demand our attention. This research talk investigates the technical roots of these issues, tracing them to biased training data and algorithmic limitations, while exploring privacy vulnerabilities stemming from model architecture, fine-tuning, and data leakage. We navigate the terrain of bias mitigation techniques, such as debiasing data and fairness-aware algorithms, and privacy-preserving methods like federated learning and differential privacy. Furthermore, we scrutinize the legal dimension, contemplating the applicability of existing regulations, AI-specific legislation, consent mechanisms, and accountability frameworks. Ethical dilemmas at the intersection of bias and privacy are discussed, highlighting trade-offs and synergies, while also envisioning a future where interdisciplinary research, industry self-regulation, and user education forge responsible AI practices that promote fairness and protect user privacy. This talk equips researchers, developers, and policymakers with a comprehensive understanding of these multifaceted challenges and the strategies to address them in large language models. Dr. Haibing Lu is full professor and department chair of Information Systems and Analytics, Leavey School of Business, Santa Clara University. He received his Ph.D. in Management with a specialization in Information Technology from Rutgers University in 2011, and Bachelor's and Master's degrees in Mathematics from Xi’an Jiaotong University, China in 2002 and 2005, respectively. Dr. Lu possesses expertise encompassing an array of domains, including data privacy and security, machine learning, AI fairness and governance, among others. His notable achievements include receiving the 2021 AACSB Innovations That Inspire distinction, the 2023 MSOM Best Paper Award, and being featured in the 2019 United Nations Human Rights Report. His scholarly contributions comprise a collection of highly-cited technical papers featured in renowned journals, including IEEE Transactions on Dependable and Secure Computing, IEEE Transactions on Intelligent Transportation Systems, IEEE Transactions on Big Data, ACM Transactions on Management Information Systems, INFORMS Journal on Computing, Manufacturing Service Operations Management, and prominent computer science conferences such as S&P, KDD, ICDM, and ICDE. Dr. Lu's research initiatives have been supported by prestigious institutions including Palo Alto Research Center, Groove, GEIRI North America, Ultimate Software, AKTANA, and Markkula Center, among others. His research findings have garnered recognition from numerous media outlets including USA Today, the United Nations, Forbes, WIRED Magazine, Yahoo Sports, AACSB, etc. reaching audiences across diverse languages. Zhongzhong Jiang Title: Content Sharing When Customers Can Multihome Abstract: In this study, we investigate the strategic decisions made by platforms regarding the sharing of exclusive content with competitors, a puzzle that has sparked considerable debate due to the observed disparity in sharing strategies across industries. For example, sharing strategy is prevalent in certain industries, e.g., video gaming, but is almost nonexistent in some other ostensibly similar industries, e.g., video streaming. Surprisingly, we identify the ratio of multihoming consumers - a factor previously overlooked - as a potential key determinant of these strategic differences. Our analysis reveals that when the ratio of multihoming consumers is low, sharing content to the competing platform leads to a win-win situation for the two platforms. Whereas when the ratio is high, sharing will not benefit both platforms. This intriguing phenomenon is primarily driven by the differing ways of valuation employed by singlehoming and multihoming consumers towards platform content. This study thus establishes a connection between customer heterogeneity in terms of platform usage and the state of content sharing among competing firms. Furthermore, we demonstrate that content sharing can lead to higher consumer surplus, highlighting the societal benefits of wide content distribution. This interesting finding has significant managerial and policy implications, which we discuss accordingly. Our study offers a fresh perspective on content sharing strategies, underscoring the pivotal role of multihoming consumers. Zhongzhong Jiang, Dean of School of Business Administration, Director of Institute of Behavioral and Service Operations Management, Northeastern University, China. He is the National “Ten Thousand People Plan” youth top talent. He was recommended to Northeastern University in 1996 and received his doctorate in systems engineering from Northeastern University in 2006. At present, he is the dean, the professor and the doctoral supervisor of the School of Business Administration, Northeastern University, the director of the Institute of Behavior and Service Operation Management, Northeastern University, the dean and the chief expert of Liaoning Service-oriented Manufacturing Research Institute, and the director of the Service-oriented Manufacturing Research Base of Liaoning Science and Technology Innovation Think Tank. He was a visiting professor at the University of Minnesota and a key member of innovation research group and international major cooperation projects of National Natural Science Foundation of China. He is also the editorial board of Production and Operations Management、Decision Sciences, and the guest editor of the International Journal of Production Research. In recent years, he has presided the National Natural Science Foundation of China focusing on the behavior and service operation management, the e-commerce and the sharing economy, logistics and supply chain optimization, etc. He has published 90 academic papers in journals such as MSOM, POM, TRB, NRL, Journal of Management Sciences in China, etc. He has obtained 2 authorized national invention patents and won 12 provincial awards such as the first prize of Outstanding achievements in Philosophy and Social Sciences of Guangdong Province, Liaoning Youth Science and Technology Award, and 4 provincial leaders’ instructions. Sihua Chen Title: When More is Less: The Other Side of Artificial Intelligence Recommendation Abstract: Based on consumers' preferences, AI (artificial intelligence) recommendation automatically filters information, which provokes scholars' debate. Supporters believe that by analyzing the consumers' preferences, AI recommendation enables consumers to choose products more quickly and with lower cost. Critics deem that consumers are more easily trapped in information cocoons because of the use of AI recommendation. This reduces the possibility of consumers contacting with a variety of commodities, thus lowering the consumer decision quality. Based on experiments, this paper discusses the moderating role of AI recommendation on the relationship of consumers' preferences and information cocoons. Moreover, it examines the relationship between information cocoons and consumer decision quality. The findings are: AI recommendation strengthens consumers' preferences; consumers' preferences are positively correlated with information cocoons and further leads to the decline of consumers’ decision quality. In the AI era, this paper contributes to revealing the dark sides of AI recommendation and provides empirical evidence for the regulation of AI behaviors. Abstract: Sihua Chen, Professor, PhD Supervisor, Director of the Research Department of Jiangxi University of Finance and Economics, Vice President of the China Information Economics Society. Hosted three National Natural Science Foundation Projects, including one Joint Key Project, one General Project, one China Postdoctoral Science Foundation Project, and more than 10 provincial and ministerial level projects. Published 2 monographs and over 50 papers. Peng Wu Title: Novel Formulations and Logic-Based Benders Decomposition for the Integrated Parallel Machine Scheduling and Location Problem Abstract: We investigate the discrete parallel machine scheduling and location (ScheLoc) problem, which consists of locating multiple machines to a set of candidate locations, assigning jobs from different locations to the located machines, and sequencing the assigned jobs. The objective is to minimize the maximum completion time of all jobs, i.e., the makespan. Though the problem is of theoretical significance with a wide range of practical applications, it has not been well studied as reported in the literature. For this problem, we first propose three new mixed-integer linear programs (MILPs) that outperform state-of-the-art formulations. Then, we develop a new logic-based Benders decomposition algorithm for practical-sized instances, which splits the problem into a master problem that determines machine locations and job assignments to machines, and a subproblem that sequences jobs on each machine. The master problem is solved by a branch-and-cut procedure that operates on a single search tree. Once an incumbent solution to the master problem is found, the subproblem is solved to generate cuts that are dynamically added to the master problem. A generic no-good cut is first proposed, which is later improved by some strengthening techniques. Two optimality cuts are also developed based on optimality conditions of the subproblem, and improved by strengthening techniques. Numerical results on small-sized instances show that the proposed formulations outperform state-of-the-art ones. Computational results on 1400 benchmark instances with up to 300 jobs, 60 machines, and 300 locations demonstrate the effectiveness and efficiency of the algorithm compared to current approaches. Peng Wu, Professor, PhD supervisor at the School of Economics and Management at Fuzhou University. He also serves as the Vice Dean of the Institute of Management Science and Engineering and as an Assistant to the Director of the Personnel Department. Dr. Wu has long been engaged in research on the theory and methods of large-scale combinatorial optimization, as well as their practical applications. He has overseen more than 10 projects, including those funded by the National Social Science Foundation and the National Natural Science Foundation of China. Furthermore, he has published over 60 papers in international journals such as the INFORMS Journal on Computing, Transportation Research Part B, Decision Support Systems, and IEEE Transactions, among others. Dr. Wu has also received recognition for his contributions, including a first prize for teaching achievements in Fujian Province, as well as two second prizes and two third prizes for outstanding achievements in the social sciences in Fujian Province. Additionally, three policy suggestion reports authored by him have been approved by the Secretary of the Fujian Provincial Party Committee and the Governor of Fujian Province. Chunwei Tian Title：An Emotional Digital Human Abstract: To capture human’s emotion and improve interaction quality between human and virtual human, an emotional digital human technique is proposed. It is composed of three parts. The first part improves a generative adverbial network to generate a digital human. Also, the digital human can change images and backgrounds, according to demands of interacted humans. The second part uses bioinformatics of faces to recognize emotions of interacted humans to enhance interaction quality. The third part utilizes a speech recognition method to achieve an interaction between a human and digital human. The emotional digital human algorithm is useful to online business of banks, medical diagnosis, entertainments, etc. Chunwei Tian received his Ph.D degree in Computer Application Technique at Harbin Institute of Technology in Jan, 2021. He was a Research Fellow at City University of Hong Kong. He is currently an Associate Professor with the School of Software, Northwestern Polytechnical University, China. Also, he is a member of National Engineering Laboratory for Integrated Aerospace Ground-Ocean Big Data Application Technology. He has become a High-level Talents of Jiangsu Province and a Young Sci-tech Talents of Suzhou Association for Science and Technology in 2022. He becomes a world’s Top 2% Scientist in 2022. He has obtained an excellent doctoral dissertation for Heilongjiang artificial intelligence society, Shenzhen CCF and Harbin Institute of Technology, respectively. His research interests include video/image restoration, image generation, digital human and deep learning. He has published over 50 papers in academic journals and conferences, including IEEE TNNLS, IEEE TMM, IEEE TSMC, Pattern Recognition, Neural Networks, Information Sciences and ICASSP, etc. He has five ESI highly-cited papers. His five paper techniques have become benchmark list for image super-resolution. Besides, he is an associate editor/young editor of the CAAI Transactions on Intelligence Technology, the Defence Technology, the Data Science and Management, etc. Feipeng Zhang Title：Efficient Nonparametric Inference for Conditional Expectile Functions with Large-Scale Time Series Data Abstract: Expectile is a coherent and elicitable law-invariant risk measure and it has been widely applied in actuarial and financial risk management. This paper focuses on developing efficient estimation and inference for nonparametric conditional expectile functions of large-scale time series data. Most existing methods based on iteratively reweighted least squares (IWLS) algorithms are not computationally feasible for nonparametric expectile estimation and inference when the sample size is big. To overcome the issue, we develop a direct nonparametric method by inverting the local polynomial estimator of the conditional loss-gain function, to estimate conditional expectile functions for time series. The proposed estimator is computationally stable with no resort to any iterative algorithms. Moreover, the proposed estimator is statistically more efficient than the existing IWLS based methods. We establish the asymptotic distributions of the proposed estimator for statistical inference of the conditional expectile functions. Monte Carlo simulations show that the proposed method is computationally efficient and has the desired finite sample performance which outperforms the expectile-loss-based method in terms of mean squared errors. We further illustrate the proposed methods using the S&P500 data set by evaluating the computational time for estimating the conditional expectile-based value-at-risk (EVaR) of S&P500 returns, and the accuracy in predicting EVaR for the out-of-sample observations. Feipeng Zhang, a professor at School of Economics and Finance, Xi’an Jiaotong University. He received a B.A. in Mathematics from Wuhan University, an M.S. in Probability from Beijing Normal University, a Ph.D in Statistics from Shanghai University of Finance and Economics. He was a postdoctoral researcher for bioinformatics at Department of Statistics, Penn State University in 2015-2018. His current research interests include financial econometrics, statistical learning, and complex data modelling, etc. He hosted three NSFC fundings and a UK Research and Innovation (UKRI) Funding. He published more than 40 papers on statistical methods in Nature Communications, Genome Research, PLOS Computational Biology, Biometrics, Science China Mathematics, etc, and on econometric applications in Technological Forecasting and Social Change, Energy Economics, International Review of Financial Analysis, etc.