General Infromation
We are proud to have confirmed two individuals who are internationally recognized for their expertise in research to be our keynote speakers. Below you may find details concerning their keynote speaches, as well as some short biographical notes on them.
Challenges in Designing Data Cyberinfrastructure Strategies
Fillia Makedon
Program Director
The Office of Cyberinfrastructure
The National Science Foundation
Fillia.S.Makedon@Dartmouth.EDU
In an age of massive data applications driving research, ubiquitous digital environments and a broader (and changing) user base, the National Science Foundation (NSF) has made it its priority to study the design and implementation of sustainable strategies for the future of Data Cyberinfrastructure (CI) in order to promote discovery and innovation. The aim is to provide stability and long-term data preservation and management of scientific data collections. Among the many challenges that span different areas of computer science, the challenge of modeling highly heterogeneous data in distributed applications is particularly suited to study by the AI research communities. In the CI context, this challenge is exacerbated by the fast that "data" can include both raw and processed data, heterogeneous and multimodal data streams, the instruments that generated these data, associated tools, methods and services, as well as the modeling of the user and the usage. AI's role is significant in this respect. In this talk, we outline a range of issues, from the capturing, modeling, storing, and preservation, to making data more easily accessible for many generations to come. We consider examples of scientific (community) collections. Viewed as "endowed collections", we explain the need for stewardship and long-term maintenance and the role of AI in providing efficient web services and distributed use of such resources.
Fillia Makedon is Professor of Computer Science at Dartmouth College since 1991 and Director of The Dartmouth Experimental Visualization Laboratory (DEVLAB) since 1992. She received her Ph.D. in Computer Science from Northwestern University in 1982. Professor Makedon is currently supervising over a dozen Ph.D. students in the areas of computational multimedia, artificial intelligence applications in cybersecurity, peer to peer systems, biomedical informatics and bioinformatics. Her team is developing datamining methods with broad impact to e-commerce, e-government, sensors, and other areas. Since Fall of 2005, she is serving a two-year term as Program Director at the National Science Foundation Office of Cyberinfrastructure where she is involved in planning strategies for the future of Data Cyberinfrastructure. She is author of numerous research articles, recipient of several NSF and other awards, recipient of a senior Fulbright award, editor of several journals and member of numerous conference committees.
Machine Analysis of Facial Expressions
Maja Pantic
Department of Computing
Imperial College London
mpantic@ieee.org
Machine understanding of facial expressions could revolutionize human-machine interaction technologies and fields as diverse as security, behavioral science, medicine, and education. In security contexts, facial expressions play a crucial role in establishing or detracting from credibility. In medicine, facial expressions are the direct means to identify when specific mental processes are occurring. In education, pupils’ facial expressions inform the teacher of the need to adjust the instructional message. In ambient intelligence and human-machine interfaces, facial expressions provide a way to communicate basic information about needs and demands to the machine. For example, certain facial signals (e.g. a wink) can be associated with certain commands (e.g. a mouse click) offering an alternative to traditional keyboard and mouse commands. The human ability to read emotions from someone’s facial expressions is the basis of facial affect processing that can lead to expanding interfaces with emotional communication and, in turn, to obtaining a more flexible, adaptable, and natural interaction between humans and machines. Consequently, computer-based recognition of facial expressions has become an active research area.
This talk reviews previous approaches to automatic facial expression analysis and provides a detailed overview of the related research efforts carried out at Delft University of Technology. It describes capabilities and limitations of current systems in sensing and analysis of emotional and social facial signals. It enumerates what we have learned using these automated systems about the spatio-temporal patterns of posed and spontaneous facial behavior and it provides an overview of conceptual and technical challenges that lay ahead.
Maja (Maya) Pantic received the MS and PhD degrees in computer science from Delft University of Technology, in 1997 and 2001. From 2001 to 2005, she was an assistant and then an associate professor at the Electrical Engineering, Mathematics and Computer Science at Delft University of Technology, The Netherlands, where she is doing research in the area of machine analysis of human interactive cues for achieving a natural, multimodal human-machine interaction (HCI). In 2006, she joined the Computing Department of the Imperial College in London, UK, where she continued her work on multimodal HCI. She is the (co-) principal investigator in two large, national, ongoing projects in the area of multimodal, affective, human-machine interaction. She was the organizer and co-organizer of various meetings and symposia on Automatic Facial Expression Analysis and Synthesis and she is the Associate Editor of the IEEE Transactions on Systems, Man, and Cybernetics - Part B: Cybernetics and the Image and Vision Computing Journal, responsible for computer vision and its applications to human-computer interaction. In 2002, for her research on Facial Information For Advanced Interface, she received Innovational Research Award of Dutch Scientific Organization as one of the 7 best young scientists in exact sciences in the Netherlands. She was an associate visiting professor at the Face Group, Robotics Institute, Carnegie Mellon University in the period June – September 2005. She has published more than 40 technical papers in the areas of machine analysis of facial expressions and emotions, artificial intelligence, and human-computer interaction and has served as an invited speaker and an organization/program committee member at several conferences in these areas. For more information, see http://mmi.tudelft.nl/~maja/
Quality Labeling of Web Content
Dr. Vangelis Karkaletsis,
Software and Knowledge Engineering Laboratory,
Institute of Informatics and Telecommunications,
National Centre for Scientific Research (N.C.S.R.) "Demokritos"
vangelis@iit.demokritos.gr
As the number of web sites and services in various domains and languages is growing rapidly, it is increasingly necessary to establish specific criteria and control measures that give consumers some guarantee that the web sites they are visiting, meet a minimum level of quality standards. Further, that the professionals offering the information or service are suitably qualified. Organisations around the world are working on establishing standards of quality in the accreditation of web content. However the establishment of codes of conduct or ethics is not enough since self-adherence to such codes is just a claim with little enforceability. It is necessary to establish rating mechanisms, either by third party accreditation or by creating thematic portals where web sites are organised and characterised against certain labelling criteria. In order for these mechanisms to be successful, they must be equipped with technologies that enable the automation of the rating process, such as information extraction techniques that allow the continuous monitoring of labelled web sites alerting the labelling agency in case some changes occur against the labelling criteria, or web crawling and spidering techniques that allow the retrieval of new unlabelled web sites, their characterisation and addition in a thematic portal.
This talk provides an overview of existing mechanisms for labeling web content and describes the use of semantic web technologies to tackle the main problem of such mechanisms, that is, the need for a continuous review and control of the accredited or filtered web sites, a process which requires a huge amount of human effort.
Dr. Vangelis Karkaletsis is a Senior Researcher at the National Center for Scientific Research (NCSR) "Demokritos" and head of the Software & Knowledge Engineering Laboratory at the Institute of Informatics & Telecommunications. Dr. Karkaletsis has substantial experience in the field of Language and Knowledge Engineering, applied to ontology engineering, web content analysis (crawling, spidering, extraction), multilingual generation, summarisation, personalization. He has coordinated or participated in several national and international RTD projects. Dr. Karkaletsis has organised or has been committee member of many workshops and conferences. He is guest Editor and reviewer in international scientific journals. He is currently Treasurer of the Greek AI Society. He recently co-founded the spin-off company 'i-sieve' technologies, which focuses on web content filtering.