Purpose The purpose of this study is to provide insights into the way in which understanding and implementation of disruptive technology, specifically big data analytics and the Internet of Things (IoT), have changed over time. The study also examines the ways in which research in supply chain and related fields differ when responding to and managing disruptive change. Design/methodology/approach This study follows a four-step systematic review process, consisting of literature collection, descriptive analysis, category selection and material evaluation. For the last stage of evaluating relevant issues and trends in the literature, the latent semantic analysis method was adopted using Leximancer, which allows more rapid, reliable and consistent content analysis. Findings The empirical analysis identified key research trends in big data analytics and IoT divided over two time-periods, in which research demonstrated steady growth by 2015 and the rapid growth was shown afterwards. The key finding of this review is that the main interest in recent big data is toward overlapping customer service, support and supply chain network, systems and performance. Major research themes in IoT moved from general supply chain and business information management to more specific context including supply chain design, model and performance. Originality/value In addition to providing more awareness of this research approach, the authors seek to identify important trends in disruptive technologies research over time.

We reveal the phenomenon that ``naive'' multivariate local polynomial regression can adapt to local smooth lower dimensional structure in the sense that it achieves the optimal convergence rate for nonparametric estimation of regression functions belonging to a Sobolev space when the predictor variables live on or close to a lower dimensional manifold.Comment: Published at http://dx.doi.org/10.1214/074921707000000148 in the IMS Lecture Notes Monograph Series (http://www.imstat.org/publications/lecnotes.htm) by the Institute of Mathematical Statistics (http://www.imstat.org

Unsupervised node embedding methods (e.g., DeepWalk, LINE, and node2vec) have attracted growing interests given their simplicity and effectiveness. However, although these methods have been proved effective in a variety of applications, none of the existing work has analyzed the robustness of them. This could be very risky if these methods are attacked by an adversarial party. In this paper, we take the task of link prediction as an example, which is one of the most fundamental problems for graph analysis, and introduce a data poisoning attack to node embedding methods. We give a complete characterization of attacker's utilities and present efficient solutions to adversarial attacks for two popular node embedding methods: DeepWalk and LINE. We evaluate our proposed attack model on multiple real-world graphs. Experimental results show that our proposed model can significantly affect the results of link prediction by slightly changing the graph structures (e.g., adding or removing a few edges). We also show that our proposed model is very general and can be transferable across different embedding methods. Finally, we conduct a case study on a coauthor network to better understand our attack method. * Work done while visiting UC Berkeley.

No abstract