To Be Confirmed

Duncan Campbell, IPTV Ltd

TBC

TBC

To Be Confirmed

Pete Steggles, Ubisense Ltd, Cambridge

Smart Space Sensors (provisional)

Pete Steggles is chief product officer of Ubisense Ltd. Ubisense arose from the demise of ATT Labs in Cambridge to develop ultra-wideband sensors, and is currently involved in many projects worldwide developing smart spaces.

To Be Confirmed

Yvonne Rogers, University of Sussex

An overview of current research

TBA

To Be Confirmed

Professor Jon Crowcroft, Computer Lab, University of Cambridge

Haggle: a Networking Architecture DesignedAround Mobile Users

Haggle is a project to build communication when there is no physical connectivity. Haggle started a bit over a year ago as a collaboration between Intel and Cambridge to look at Opportunistic Networking - this is a variant of Delay or Disruption Tolerant Networking, initially devised by Vint Cerf and Kevin Fall as an architecture for Interplanetary communication for NASA's missions to Mars and other planets. Due to limitations imposed by the speed of light, round trip times are on the order of 10s of minutes. Due to conditions imposed by the Sun, interference can be quite severe lasting minutes hours or even days. Mars may be occluded by other small objects such as the moon.

In Haggle, we have the more modest goal of providing human communication using gadgets that people carry. The use of mobility, the physical movement of objects in space (those gadgets carried by people, letters, planes, trains and automobilies) is treated as part of the network, just as bits going down wires, or over the airwaves.

The applications need to be fully decoupled. Indeed, in Haggle we do not think in terms of senders and recipients at all - data is either desired or offered. Content is a bit more subtle than in layered networks (Haggle is not conventionally layered either:) so "Content Based Addressing" is not quite the right mental model of what we are up to.

Early results include: detailed measurement of human mobility; mathematical analysis of potential classes of forwarding/dissemination algorithms; prototype architecture and application implementation; a number of user trials.

Haggle is now a full Future and Emerging Technologies (FET) Integrated Project funded under the Situated and Autonomic Communication program of the Information Society Technologies priority area of the European Union's Framework Programme 6 (FP6). Work under Haggle is commenced in January 2006 and the project will last 4 years.

The name? oh, Christophe Diot (l'diot) and I were discussing Ad Hoc networking - I decided that what was needed in disconnected networks was the moral equivalent of ad hoc google, applied to _everything (name/address services, location, content, everything) so I called it Ad Hoc Google (Haggle also has connotations of negotiation - the resource management model is based on trust and recommendation networks woth agents acting on behalf of users - see?). Christophe said "Had Oc Google", and the name stuck. Those of you who've read about the sad history of Cathars and the radical Albigensien Creed in 13th centuary in the part of France where they spoke Langue D'Oc will also appreciate the multidimensional nature of the name. (Also Oc == Occidental...).

This talk will concentrate on the system architecture - papers are avaialble from the project pages: http://www.cambridge.intel-research.net/haggle/publications.php (or you can google for them - next yr, haggle for them:)

Mon May 15 17:00:00 2006

Rolf Pfeifer, Artificial Intelligence Laboratory, University of Zurich, Switzerland

Morphological computation - connecting brain, body, and environment

Traditionally, in robotics, artificial intelligence, and neuroscience, there has been a focus on the study of the control or the neural system itself. Recently there has been an increasing interest into the notion of embodiment in all disciplines dealing with intelligent behavior, including psychology, philosophy, and linguistics. In this talk, we explore the far-reaching and often surprising implications of this concept. While embodiment has often been used in its trivial meaning, i.e. .intelligence requires a body., there are deeper and more important consequences, concerned with connecting brain, body, and environment, or more generally with the relation between physical and information (neural, control) processes. Often, morphology and materials can take over some of the functions normally attributed to control, a phenomenon called .morphological computation.. It can be shown that through the embodied interaction with the environment, in particular through sensory-motor coordination, information structure is induced in the sensory data, thus facilitating perception and learning. A number of case studies are presented to illustrate the concepts introduced. I conclude with some speculations about potential lessons for robotics.

Bio: Rolf Pfeifer received his masters degree in physics and mathematics and his Ph.D. in computer science from the Swiss Federal Institute of Technology (ETH) in Zurich, Switzerland. He spent three years as a post-doctoral fellow at Carnegie-Mellon University and at Yale University. Since 1987 he has been a professor of computer science at the Department of Informatics, University of Zurich, and director of the Artificial Intelligence Laboratory. Having worked as a visiting professor and research fellow at the Free University of Brussels, the MIT Artificial Intelligence Laboratory, the Neurosciences Institute (NSI) in San Diego, and the Sony Computer Science Laboratory in Paris, he was elected "21st Century COE Professor, Information Science and Technology" at the University of Tokyo for 2003/2004, from where he held the first global, fully interactive, videoconferencing-based lecture series "The AI Lectures from Tokyo" (including Tokyo, Beijing, Jeddah, Warsaw, Munich, and Zurich). His research interests are in the areas of embodiment, biorobotics, artificial evolution and morphogenesis, self-reconfiguration and self-repair, and educational technology. He is the author of the book "Understanding Intelligence", MIT Press, 1999 (with C. Scheier). His new popular science book entitled "How the body shapes the way we think: a new view of intelligence," MIT Press, 2006 (with Josh Bongard) is scheduled to appear this summer.

Mon Mar 27 16:00:00 2006

Professor Gordon McCalla, University of Saskatchewan, Canada

The Ecological Approach to E-Learning

After some preliminary discussion of the area of artificial intelligence in education (AIED) as a "crucible for AI research", this talk will then focus on the ecological approach to e-learning. In the ecological approach learning material is assumed to be kept in repositories of learning objects that learners interact with through the mediation of an AIED system. Data is kept about each learner and each learner's interactions with each learning object. Over time this data builds up and can be "mined" to find patterns of interaction that can inform the AIED system and allow it to adapt to individual learner differences. The ecological approach also promotes the notion that both learners and learning objects are represented by agents who can negotiate learning goals and other aspects of mutual interest. This approach is currently being explored in the ARIES Laboratory in research projects investigating the I-Help peer help system, exploring active learner modelling, building a research paper recommender, and designing adaptive systems as part of the LORNET Canadian research network.

Brief Biography: Gord McCalla is a Professor in the Department of Computer Science at the University of Saskatchewan in Saskatoon, Canada. His research interests are in applied artificial intelligence, focussed particularly on user modelling and artificial intelligence in education (AIED). Working with colleagues and students in the ARIES Laboratory at the U. of S., Gord has explored many issues, including granularity in learning and reasoning, educational diagnosis, learner modelling, tutorial dialogue, instructional planning, peer help, and learning object repositories. Recently, he has begun to look into the implications of "fragmented learning systems", systems that are designed to support learners in diverse virtual learning communities (social fragmentation) and systems that are themselves composed of many software agents (technological fragmentation). This has led to an AIED architecture called the "ecological approach", currently being explored in a number of research projects in the ARIES Laboratory. Gord is a former President of the International AIED Society and is currently President of the Canadian Association of Computer Science, the organization of Canadian University Computer Science Departments.

Wed Mar 8 16:00:00 2006

Dr Andrew Moore, Queen Mary, University of London

Towards Accurate Network-Traffic Characterization

Accurate traffic classification is the keystone of numerous other network activities, from security monitoring to accounting, and from Quality of Service to providing operators with useful forecasts for long-term provisioning. Well-known port numbers can no longer be used to reliably identify network applications. There is a variety of new Internet applications that either do not use well-known port numbers or use other protocols, such as HTTP, as wrappers in order to go through firewalls without being blocked. One consequence of this is that a simple inspection of the port numbers used by flows may lead to the inaccurate classification of network traffic.

With a motivation to provide accurate traffic characterization this talk will cover issues of network monitoring, the challenges of traffic characterization and discuss some results gained using both labour-intensive and more broadly-applicable techniques.

Biography: Andrew Moore has recently joined Queen Mary, University of London after more than 10 years in Cambridge. Working with the computer science, electrical engineering and mathematics departments he leads a research group investigating topics of computer networking, including monitoring, characterization and performance analysis.

Mon Mar 6 16:00:00 2006

Murray Shanahan, Imperial College, London

A Computational Model of Conscious Information Processing

This talk describes a brain-inspired cognitive architecture that incorporates approximations to the concepts of consciousness, imagination, and emotion. To emulate the empirically established cognitive efficacy of conscious as opposed to non-conscious information processing in the mammalian brain, the architecture adopts a model of information flow from global workspace theory. Cognitive functions such as anticipation and planning are realised through internal simulation of interaction with the environment. Action selection, in both actual and internally simulated interaction with the environment, is mediated by affect. An implementation of the architecture is described which is based on weightless neurons and is used to control a simulated robot.

Mon Dec 5 16:00:00 2005

Richard Cox, Representation and Cognition Group, University of Sussex

An overview of current research in the Applied Cognition lab

In the talk I will provide an overview of some topics of current research by members of the Applied Cognition lab. These include clinical reasoning, vicarious learning, information display selection and `design for attention'. Three projects will be described in more detail: 1. an ESRC-TLRP funded project that I direct which is called `Vicarious learning and case-based teaching of clinical reasoning skills'; 2. work by my PhD student Beate Grawemeyer on an 'AIVE' - an adaptive system that assists users to select appropriate information visualisations and 3. some recent collaborative work with colleagues in the Representation and Applied Cognition research group on 'Design for Attention' in which we explored the implications of recent human attention research findings for the design of attention-aware systems.

Mon Nov 28 16:00:00 2005

David Young, Vision Group, University of Sussex

Computer Vision: from seeing to looking?

The most important trend in recent Computer Vision research has been from passive seeing to active looking: rather than analysing given images, seeking out useful information from the optic array. This talk is about aspects of research that reflect this change: foveal vision, using models of the primate retina; saccadic camera movements, using biologically-inspired mechanisms; local patterns of optic flow for the control of action. I will also say a little about developing statistical models to support the higher levels of active vision.

Mon Nov 21 16:00:00 2005

Inman Harvey, EASY Group, University of Sussex

EASy and the CCNR, research in the Evolutionary and Adaptive Systems group

The EASy group is the world's largest research group on the interface between computer science and AI, on the one hand, and biological ideas and methods on the other hand. The CCNR is our research centre, shared between Informatics and Life Sciences. I will give a brief overview of the last 15 years, and of current research.

Mon Apr 25 16:00:00 2005

Alan Blackwell, Cambridge University

Interdisciplinary adventures in abstraction and representation

This talk presents research results from the Crucible network for research in interdisciplinary design, and the Vital Signs project investigating future notations and representational systems. In many domains of human life, the spread of digital devices requires people to think about their activities in more abstract ways. People work with abstractions not directly, but by using representations and notations. In many cases, they become "end-user programmers", whether they want to or not. The Crucible and Vital Signs teams have been investigating this trend from ethnographic, cognitive and design perspectives over the past five years. Alan Blackwell will summarise results so far with reference to several different application domains, including home automation, web queries, pedagogical rhetoric, and more powerful spreadsheets. http://www.cl.cam.ac.uk/~afb21/vital/ http://www.crucible.cl.cam.ac.uk/

Mon Mar 21 16:00:00 2005

Luca Cardelli, Microsoft Labs, Cambridge

Abstract Machines of Systems Biology

Living cells are extremely well-organized autonomous systems, consisting of discrete interacting components. Key to understanding and modeling their behavior is modeling their system organization. Four distinct chemical toolkits (classes of macromolecules) have been characterized, each combinatorial in nature. Each toolkit consists of a small number of simple components that are assembled (polymerized) into complex structures that interact in rich ways. Each toolkit abstracts away from chemistry; it embodies an abstract machine with its own instruction set and its own peculiar interaction model. These interaction models are highly effective, but are not ones commonly used in computing or concurrency theory (or mathematics): proteins stick together, genes have fixed output, membranes carry activity on their surfaces. "Systems biology" consists, largely, in understanding how these interaction models work, separately and together. To that end, biologists have invented a number of notations attempting to describe, abstractly, these abstract machines and the processes and networks they implement. I discuss the notations currently used by biologists, and the advantages of using programming language (process calculus) approaches. The long-term goal is to represent the structure and function of biological systems via formal languages, for description, simulation, analysis and (eventually) synthesis.

Mon Feb 28 16:00:00 2005

David Duce, Oxford-Brookes University

Collaborative Visualization in Grid Computing Environments

This talk will give an overview of distributed and collaborative visualization and then describe some recent developments. We have devised a model for dataflow visualization at three levels: an abstract specification of the intent of a visualization, a binding of these abstract modules to a specific software system; and then a binding of software to processing and other resources. We have an XML application (language), called skML, capable of describing visualization at the three levels. A feature of the approach is that a visualization can first be described using modules from a generic collection and then translated into a concrete collection of modules in a specific visualization system, drawing on a variety of Web technologies to do so. We are currently developing an application scenario based on wildfire management. We are developing a new collaborative visualization tool (using Web technology) within the context of an architectural framework for Grid middleware being developed at the University of Lancaster. The talk will outline the approach we are taking and the thinking behind it. Through this work we have found the need for an ontology of visualization. The talk will conclude with some thoughts about this. The talk is based on joint work with the University of Leeds in the gViz project (under the e-Science Core Programme) and the University of Lancaster in the Open Overlays project (under the e-Science Fundamental Computer Science for e-Science Programme).

Mon Jan 31 16:00:00 2005

Ian Horrocks, University of Manchester

Reasoning with Expressive Description Logics: Logical Foundations for the Semantic Web

Description Logics (DLs) are a family of logic based Knowledge Representation formalisms descended from semantic networks and KL-ONE. They are distinguished by having formal (model theoretic) semantics, and by the provision of (sound and complete) inference services, with several highly optimised implementations now being available. DLs have a wide range of applications, but are perhaps best know as ontology languages (they provide the basis for recent "Semantic Web" ontology languages such as OIL, DAML+OIL and OWL). In this talk I will give a brief history of DLs and of DL applications, in particular their application in the context of the Semantic Web. If time permits, I will then give an overview of the reasoning techniques that are employed by state of the art DL implementations, and which enable them to be effective in realistic applications, in spite of the high worst case complexity of their basic inference problems. Finally, I will point out some interesting areas for future research, in particular those related to the Semantic Web application area.

Mon Nov 29 16:00:00 2004

Ben Varcoe, University of Sussex

Cancelled: Quantum Computing

CANCELLED

Mon Nov 22 16:00:00 2004

Timothy Roscoe, Intel Labs, Berkeley

PlanetLab Design Principles and Experiences

PlanetLab is a geographically distributed platform for deploying, evaluating, and accessing planetary-scale network services. PlanetLab is a shared community effort by a large international group of researchers, each of whom gets access to one or more isolated "slices" of PlanetLab's global resources via a concept we call distributed virtualization. In order to encourage innovation in infrastructure, PlanetLab decouples the operating system running on each node from a set of multiple, possibly 3rd-party network-wide services that define PlanetLab, a principle we refer to as unbundled management. This talk will examine the design principles of PlanetLab together with the experience we have gained in building and operating the platform over the past two years.

Mon Nov 15 16:00:00 2004

Maggie Boden, University of Sussex

The Mathematical Biology of D'Arcy Thompson

D'Arcy Thompson (1860-1948) was a maverick biologist who raised many intriguing questions, and suggested many imaginative answers, that still feel 'modern' nearly a century later. (One might also say that they feel over two hundred years old: his ideas on biological form were much influenced by Johann von Goethe, who was highly unfashionable in D'Arcyh's day--and ours.) His 1917 book "On Growth and Form" enthused many people at the time. But his ideas couldn't be followed up. The pioneering mathematical biology he outlined couldn't be taken much further without the help of computers. He was a 'grandfather' of A-Life, and would have been an A-Lifer today.