Keynote 1: Interaction-Oriented Programming for Decentralized Service Engagements

Munindar P. Singh, North Carolina State University

Abstract

This talk introduces the conception of Interaction-Oriented Programming (IOP), a way of organizing distributed systems comprising autonomous and heterogeneous components. Crucially, IOP gives first-class status to interactions as opposed to objects or agents.

IOP begins from a fully declarative approach to interaction protocols an architectural abstractions of service engagements that capture the operational aspects of interactions purely in terms of causality and integrity properties and eliminates hidden control states and dependencies. The resulting model, first, enables capturing application meaning declaratively in terms of commitments and other abstractions for business contracts. In this way, it lends itself to flexible service engagements, yet enables formally verification for liveness and safety. Second, the model is naturally realized in a shared-nothing architecture that decouples service endpoints except as necessary for capturing application meaning. Third, it enables application-level fault tolerance wherein the endpoints reason about their expectations based on a protocol. Fourth, IOP is naturally realized in modern data-driven architectures, including serverless computing.

Speaker Bio

Dr. Munindar P. Singh is an Alumni Distinguished Graduate Professor in the Department of Computer Science at North Carolina State University. Munindar's research interests include artificial intelligence and multiagent systems with applications in service-oriented computing, cybersecurity, privacy, and social computing. He is a codirector of the DoD-sponsored Science of Security Lablet at NCSU, one of six nationwide.

Munindar was the editor-in-chief of the ACM Transactions on Internet Technology from 2012 to 2018 and the editor-in-chief of IEEE Internet Computing from 1999 to 2002. His current editorial service includes IEEE Internet Computing, Journal of Autonomous Agents and Multiagent Systems, IEEE Transactions on Services Computing, and ACM Transactions on Intelligent Systems and Technology. His previous editorial service includes the Journal of Artificial Intelligence Research and the Journal of Web Semantics. Munindar served on the founding board of directors of IFAAMAS, the International Foundation for Autonomous Agents and MultiAgent Systems. He also served on the founding steering committee for the IEEE Transactions on Mobile Computing. Munindar was a general cochair for the 2005 International Conference on Autonomous Agents and MultiAgent Systems and a general cochair for the 2016 International Conference on Service-Oriented Computing.

Munindar is a Fellow of the Institute of Electrical and Electronics Engineers (IEEE), the Association for the Advancement of Artificial Intelligence (AAAI), and the American Association for the Advancement of Science (AAAS), and a member of Academia Europaea. He has won the ACM/SIGAI Autonomous Agents Research Award, the IEEE TCSVC Research Innovation Award, and the IFAAMAS Influential Paper Award. He won NC State University's Outstanding Research Achievement Award twice, was selected as an Alumni Distinguished Graduate Professor, and is a member of NCSU's Research Leadership Academy. He won NCSU's Faculty Graduate Mentor Award.

Munindar's research has been recognized with awards and sponsorship by (alphabetically) Army Research Lab, Army Research Office, Cisco Systems, Consortium for Ocean Leadership, DARPA, Department of Defense, Ericsson, Facebook, IBM, Intel, National Science Foundation, and Xerox. Twenty-nine students have received PhD degrees and thirty-nine students MS degrees under Munindar's direction.

Keynote 2: Fuzzy Transfer Learning

Jie Lu, University of Technology Sydney

Abstract

This talk will describe how fuzzy transfer learning can innovatively and effectively learn from data to support data-driven decision-making in uncertain and dynamic situations. The core idea behind fuzzy transfer learning is to leverage previously acquired knowledge to assist in completing a prediction task in a related domain by integrating fuzzy techniques with the transfer learning process. A set of new fuzzy transfer learning theories, methodologies, and algorithms is introduced, which transfers knowledge learned in one or more source domains to target domains. The fuzzy transfer learning set incorporates (1) a fuzzy refinement domain adaptation algorithm by utilizing the fuzzy system and similarity/dissimilarity concepts to modify the target instances' labels for classification; (2) fuzzy rule-based systems with mapping functions by building latent spaces to facilitate knowledge transfer for regression tasks in both homogeneous and heterogeneous scenarios; (3) unsupervised domain adaptation, to recognize newly emerged patterns in target domains that may be unlabelled. Patterns in target domains are recognized by leveraging knowledge from patterns learned from source domains and solutions to heterogeneous unsupervised domain adaptation via n-dimensional fuzzy geometry and fuzzy equivalence relations. These new developments can enhance data-driven prediction and decision support systems in complex real-world environments.

Speaker Bio

Distinguished Professor Jie Lu is a scientist in the field of computational intelligence, primarily known for her work in fuzzy transfer learning, concept drift, recommender systems, and decision support systems. She is an IEEE Fellow, IFSA Fellow, and Australian Laureate Fellow. Currently, Prof Lu is the Director of the Australian Artificial Intelligence Institute (AAII) and Associate Dean (Research Excellence) at the Faculty of Engineering and Information Technology, University of Technology Sydney (UTS). She has published over 400 papers in leading journals and conferences; won 10 Australian Research Council (ARC) Discovery Projects and 15 industry projects; and supervised 46 doctoral students to completion. Prof Lu serves as Editor-In-Chief for Knowledge-Based Systems and International Journal of Computational Intelligence Systems, and is a recognized keynote speaker, delivering 30 keynote speeches at international conferences. She is the recipient of the IEEE Transactions on Fuzzy Systems Outstanding Paper Award (2019), the Computer Journal Wilkes Award (2018), Australia's Most Innovative Engineer Award (2019), and the UTS Chancellor's Medal for Research Excellence (2019).

Keynote 3: Privacy in the AI-Enabled World

James B. D. Joshi, University of Pittsburgh

Abstract

Recent advances in computing and information technologies has enabled a hyper-connected cyberspace that has become an intricate part of our society. Enabled by such connectivity and the growing computational power/infrastructures at our disposal, innovative AI and Machine Learning (ML) techniques are increasingly being deployed in various applications. Such AI-driven world has been further fueled by huge amounts of data that is being continuously collected in myriad of ways, including data that has or can reveal highly privacy-sensitive information about us. While AI/ML technologies and the huge amounts of data available can be used for immense benefits for our society, privacy issues pose as a roadblock. Increasing number of privacy regulations being introduced to protect privacy as a basic right further brings various challenges in terms of use of data by AI/ML-enabled applications. In this talk, I will discuss current state of privacy mainly within the context of AI/ML, including current state of privacy-preserving technologies, and discuss various challenges and potential research directions.

Speaker Bio

James Joshi is a professor of School of Computing and Information at the University of Pittsburgh, and the director/founder of the Laboratory of Education and Research on Security Assured Information Systems (LERSAIS). He is currently serving as an NSF Program Director in the Computer and Network System (CNS) division, and in the Secure and Trustworthy Cyberspace (SaTC) program. He currently also serves as the Co-Chair of the Privacy Interagency Working Group of Networking and Information Technology R&D (NITRD), which coordinates federal R&D. He is an elected Fellow of the Society of Information Reuse and Integration (SIRI), a Senior member of the IEEE and a Distinguished Member of the ACM. His research interests include access control models, security and privacy of distributed systems and AI/ML, and trust management. He is a recipient of the NSF CAREER award in 2006. He established and managed the NSF CyberCorp Scholarship for Service program at Pitt in 2006 (two rounds). He also established LERSAIS as a NSA designated Center of Academic Excellence in Cyber Defense (both CAE and and CAE-R). He has served as a program co-chair and/or general co-chair of several international conferences/workshops, including, ACM SACMAT, IEEE BigData, IEEE IRI, IEEE CIC, IEEE ISM, IEEE EDGE, etc. He currently serves as the steering committee chair of IEEE CIC/TPS/CogMI, as the EIC of IEEE Transactions on Services Computing. He has published over 140 articles as book chapters and papers in journals, conferences and workshops, and has served as a special issue editor of several journals including Elsevier Computer & Security, ACM TOPS, Springer MONET, IJCIS, and Information Systems Frontiers. His research has been supported by NSF, NSA/DoD, and Cisco.

Keynote 4: Efficient Network Embeddings for Large Graphs

Xiaokui Xiao, National University of Singapore

Abstract

Given a graph G, network embedding maps each node in G into a compact, fixed-dimensional feature vector, which can be used in downstream machine learning tasks. Most of the existing methods for network embedding fail to scale to large graphs with millions of nodes, as they either incur significant computation cost or generate low-quality embeddings on such graphs. In this talk, we will present two efficient network embedding algorithms for large graphs with and without node attributes, respectively. The basic idea is to first model the affinity between nodes (or between nodes and attributes) based on random walks, and then factorize the affinity matrix to derive the embeddings. The main challenges that we address include (i) the choice of the affinity measure and (ii) the reduction of space and time overheads entailed by the construction and factorization of the affinity matrix. Extensive experiments on large graphs demonstrate that our algorithms outperform the existing methods in terms of both embedding quality and efficiency.

Speaker Bio

Xiaokui Xiao is a Dean's Chair Associate Professor at the School of Computing, National University of Singapore (NUS). His research focuses on data management, with special interests in data privacy and algorithms for large data. He received a Ph.D. in Computer Science from the Chinese University of Hong Kong. Before joining NUS, he was an associate professor at the Nanyang Technological University, Singapore. He received the best research paper award in VLDB 2021, and is serving as an associate editor for the VLDB Journal and the IEEE Transactions on Knowledge and Data Engineering.