Northrop Grumman Corporation, Virginia, USA
Title: Integrating Semantic Computing Capabilities into Mission Critical Systems
Abstract: Large scale heterogeneous mission systems are composed of a wide range of components, software, hardware and architectures that are integrated across a variety of platforms and frameworks. Semantic computing provides a practical transition path from complex system automation through cognitive processing and towards artificial intelligence, allowing us to achieve optimum performance and enhanced effectiveness of mission capabilities. Considering the multi-faceted dimensions of Northrop Grumman’s mission objectives, an end-to-end systems-level design approach is required to leverage and integrate capabilities across the enterprise and achieve an integrated semantics system-of-systems, at scale, in tactical and strategic environments. The presentation will demonstrate various examples in which semantic processing applies to Northrop Grumman systems, products and services. The discussion will also provide insights into practical limitations of semantic computing used within mission critical applications.
Bio: Mr. Paul Conoval is Director, Technology for Northrop Grumman’s Mission Systems (MS) Sector. In this role he leads a wide range of advanced technology initiatives including the integration of the sector’s R&D program, university research, intellectual property, innovation, and development of emerging technologies applied to Intelligence and DoD. His prior work includes development of national-scale intelligence, SIGINT and communications systems with emphasis in information theory, digital signal processing and electronics hardware design. Prior to 2009, Mr. Conoval served as Chief Technology Officer of TASC Inc. in Chantilly, VA, and also worked at MITRE Corporation and the Singer Corporation Kearfott Division developing digital avionics and communications systems. Mr. Conoval holds a BSEE from the Cooper Union School of Engineering and an MSEE from Rutgers University. Mr. Conoval is also an active practitioner of intellectual property and as a registered patent agent of the US Patent and Trademark Office.
Chris Dede and Joseph Reily
Title: Scaffolding Sophisticated Performances: Interpreting Steams of Learner Behaviors in Open-Ended Complex Settings
Abstract: Well established strategies are available for analyzing student's micro-actions in closed-ended learning experiences, such as tutoring systems, and research has shown the advantages of the real-time feedback this can provide. However, these methods degrade in open-ended learning experiences, which can teach a wider range of knowledge and skills, because interpreting students' behaviors is much more complex. As an example of an approach to this challenge, our EcoLENS project is using streams of logfile data to scaffold learning in inquiry-based open-ended virtual environments. A constraint is that this feedback must not over-structure their experience, as this will undercut the benefits of the freedom they are afforded. Our analyses thus far have applied several different approaches toward detecting different types of student behaviors and reporting them back to the student or teacher in meaningful ways. How groups use their time in the virtual world can be visualized and reported to the classroom teacher to identify outliers and direct teacher attention. Specific patterns or sequences of logged actions can be mined and associated with positive or negative behaviors in the world such as searching for disconfirming evidence, being stuck, or testing a hypothesis. Within certain repeated activities such as virtual experiments, groups' problem solving strategies can be deduced and tracked over time to see how they evolve and mature. Groups' logged actions in the virtual world can be filtered to provide a time series trajectory showing the rate of their investigative activities over time, looking at the entire course of their experience instead of specific subsequences. Trajectories can then be grouped via clustering to identify different typical pathways which can be correlated with learning gains. Methodological and instructional design advances of this type can aid semantic computing situations in which learners are experiencing immersive authentic simulations.
Bios: Chris Dede is the Timothy E. Wirth Professor in Learning Technologies at Harvard's Graduate School of Education (HGSE). His fields of scholarship include emerging technologies, policy, and leadership. In 2007, he was honored by Harvard University as an outstanding teacher, and in 2011 he was named a Fellow of the American Educational Research Association. From 2014-2015, he was a Visiting Expert at the National Science Foundation Directorate of Education and Human Resources. His edited books include: Scaling Up Success: Lessons Learned from Technology-based Educational Improvement, Digital Teaching Platforms: Customizing Classroom Learning for Each Student, Teacher Learning in the Digital Age: Online Professional Development in STEM Education, Virtual, Augmented, and Mixed Realities in Education, and Education at Scale: Engineering Online Learning and Teaching.
Joseph Reilly is an advanced doctoral student at the Harvard Graduate School of Education where he conducts data-intensive research in education with the EcoLearn team and the LIT Lab. He holds a B.S. in Chemistry from Georgetown University and a M.A. in Special Education from American University. Joe taught middle school science in Washington D.C. and high school chemistry in Virginia prior to beginning his doctoral work. His research interests include immersive virtual environments for learning, multi-modal learning analytics, stealth assessment of student understanding, automatic generation of dynamic scaffolding, and leveraging data science techniques in educational research.
Thursday 01/31 (Banquet)
Vrije Universiteit Amsterdam & CWI, The Netherland
Bio: Dick Bulterman is a member of the faculty of the department of computer science at the Vrije Universiteit (VU) in Amsterdam, The Netherlands. He is also a Fellow at Centrum Wiskunde en Informatica (CWI), also in Amsterdam, where he has worked on multimedia systems, languages and user interfaces for more than thirty years. Prof. Bulterman was President and CEO of the FX Palo Alto Laboratory (FXPAL) in California from 2013-2016 and CEO of Oratrix Development from 1998-2002.
Bulterman’s main research interests include the development of declarative multimedia languages (leading to the development of W3C’s SMIL standard for temporal control of XML content) and the development of user agents that can flexibly and interactively control the content of media presentation in a distributed context (leading to the GRiNS and Ambulant multimedia engines). He has contributed to a dozen European and national research projects, has provided fundamental contributions to hypermedia languages and has studied user behavior in creating and experiencing media content.
In addition to his work at the VU and CWI, Bulterman is chair of ACM SIGWEB. He received the ACM SIGMM Lifetime Achievement Award in 2013. He was a long-time editorial board member of ACM Transactions on Multimedia, Multimedia Systems and Multimedia Tools and Applications. He has been an organizer of conferences for major ACM and IEEE conferences in the area of multimedia and multimedia systems. He received his PhD from Brown University in Providence, RI, initially specializing in computer graphics architectures.
Ford Motor Company
Title: Scalable Analytics for Emerging Applications in the era of Industry X.0
Abstract: Given the commoditization of analytics methods and the avalanche of Industry X.0 data flowing from customer interaction and life-cycle functioning of products, building a proof-of-concept (PoC) data and analytics pipeline for customer-facing emerging application has become trivial. However, to truly leverage such an endeavor to create business value in a seamless friction-free manner requires scaling up the analytics in a nontrivial way. This presentation will focus on identifying some of inherent challenges in scaling up analytics in a legacy environment. We will discuss some of the Scalable Analytics systems that the presenter has been involved in throughout his career at Ford and demonstrate how these systems are deployed in industry.
Bio: Amit Mohanty leads the Connected Vehicle Analytics group as part of Global Data, Insight and Analytics at Ford Motor Company. His team is responsible for developing advanced analytics and machine learning algorithms for Ford's connected vehicle data. Previously, he was the founding engineer for controls software group at a Vinod Khosla start-up - EcoMotors, Inc. Amit began his career as a staff scientist with US Department of Energy. He received his Ph.D. from Purdue University, M.S. from Southern Illinois University and B.Tech. from Indian Institute of Technology, Kharagpur - all in Mechanical Engineering.