ECS Intranet:
Decentralised Information Systems
The growth of the web presents users with a means of accessing an ever increasing number of diverse information sources (ranging from institutional archives to real world deployed sensor networks). However, using of this information to make informed decisions presents a number of challenges. The information may be incomplete, uncertain or contradictory. It may come from sources owned by different stakeholders, and increasingly, it may incorporate a huge number of different media. Thus, there is a clear need for systems that are able to interact with these diverse information sources to not only collect relevant information, but also to reformulate and reason about it in principled ways.
To this end, agent-based computing has been identified as an appropriate engineering paradigm for building such decentralised information systems. Within IAM we are designing, prototyping and evaluating the technologies that are required to build these systems. These technologies range from negotiation techniques to allow agents to cooperate to form ad-hoc networks that exchange relevant information, recommender systems to assist users in finding information relevant to their current context, and web technologies that promote the inter-operability of information systems.
Research Areas
Decentralised Agent Systems
In order to deploy agent-based solutions we are investigating the fundamental issue of how agents should communicate, negotiate and cooperate. This work explores novel paradigms such as market-based control and combines techniques from game theory, economics and evolutionary science to engineer robust systems that operate under a decentralised control regime.
Ubiquitous and Mobile Agents
As sensor networks and mobile computers become ubiquitous within the environment, we expect these devices to form an essential part of future information systems. Thus we are exploring methodologies that allow huge numbers of both stationary and mobile agents to seamlessly interact with one another.
Web Technologies
To enable the exchange of information between diverse sources, we are exploring and developing web technologies that support and promote inter-operability. Examples of these include hypermedia linking to maintain first-class associations between documents used to access to information, the use of metadata to describe and classify multi-media content, and the design and maintenance of ontologies that define and capture application specific terms and knowledge.
Projects
- Adaptive Energy-Aware Sensor Networks
- Agent-Based Control of Decentralised Systems
- AIMS: Academic Information Management Systems
- ALADDIN: Autonomous Learning Agents for Decentralised Data and Information Networks
- ARGUS II: Decentralised Data Fusion
- Autonomic Supply Chains in Computational Economies
- CogPrints: Cognitive Sciences EPrint Archive
- COHSE: Informed WWW Link Navigation Using Ontologies
- Communities of Collocation
- Control and Management of Autonomous Mobile Sensors
- e-Stat
- EnAKTing the Unbounded Data Web: Challenges in Web Science
- Equator
- FEEL: Non-intrusive services to support focused, efficient and enjoyable local activities
- Intelligent Decentralised Energy-Aware Systems (iDEaS)
- Knowledge Capture, Sharing and Reuse in the Design Process
- LinkMe: Distributed Link Services for Mobile Environments
- Magnitude: Mobile AGents Negotiating for ITinerant Users in the Distributed
- Market-Based Control of Complex Computational Systems
- Market-Based Recommender Systems
- MOHICAN: MObile Handsets In Cooperative Agents Network
- Ontoportal
- OpCit: The Open Citation Project
- OpenKnowledge
- PASOA: Provenance-Aware Service-Oriented Architecture
- QuIC: Queries in Context