Professor Wendy Hall

Home | Research Overview


Research Overview

From the mid 1980s I have been interested in hypermedia and multimedia, and developing ways to enable people to interact with, and infer knowledge from, large-scale multimedia information systems. I started work on the Microcosm hypermedia system in 1989. Coincidentally this was the same year that Tim Berners-Lee first proposed the development of the system that became known as the World Wide Web (WWW) to the management at CERN. The significance of the research embodied in Microcosm is now becoming more apparent as the Web community develops the Semantic Web (the next generation of the Web). Many of the ideas we pioneered in Microcosm are now encapsulated in the key elements of the Semantic Web, in particular the dynamic linking of information based on content, context and metadata.

Whilst the Web provided the standards and protocols to enable the interconnection of documents in a global network, the philosophy and design of Microcosm [Hall et al, 1996] focussed on the nature of those interconnections (or links). The architecture was based on a set of communicating filters that monitored the messages flowing between the user and the system and acted upon them to add or filter out information as appropriate to the user. It therefore provided a framework for developing adaptive and customised hypermedia systems. The separation of links and content enabled different sets of links to be provided for different users in different contexts. The links were created dynamically on the basis of the content and the context of the information, and effectively created a metadata layer that described the information they pointed to, all ideas that are now being developed as part of the Semantic Web.

The basic tenets that drove the design of Microcosm in the early 1990’s were equally applicable in the then new Web environment. We developed several versions of a web-based Distributed Link Service (DLS) to experiment with the management of link information separately from documents [De Roure et al 1996, Carr et al, 1998]. This work is cited as a key influence on the development and design of the XLink protocol, which is part of the XML (eXtended Markup Language) standard for generating Web documents developed by the Web community to replace HTML.

By a “link service”, we mean a network service that encapsulates responsibility for link resolution and maintenance. As described in Hall [2000b] a content-based linking approach considerably reduces the authoring effort in Web applications. In the commercial version of the DLS, developed by Active Navigation Ltd the company that span out of our research, the associative links are created using summarising software that creates a theme index for the document set and uses this to create the linkbase(s). In the research version, we experimented with various methods of automatically generating linkbases, including the use of an ontology server (Carr et al, 2001). The use of ontologies to create links in the link service was very much a precursor for the current work in the Web community to develop the Semantic Web. In the MAVIS projects we developed these same ideas to apply content and context based linking to non-text media such as images and video [Lewis et al 1996] and went on to develop a multimedia thesaurus approach to this work which again foresaw Semantic Web developments [Lewis et al, 1998].

In Microcosm, as mentioned previously, one of the key ideas was to generate personalised or customised information environments using a series of filters. In the 1990’s this part of the Microcosm architecture evolved to make use of agent-based technologies. Put simply an agent is a computer process that acts autonomously to perform a particular task, or set of tasks, on behalf of the user. Agent-based technologies facilitate the development of modular, flexible, distributed systems and also help to overcome the limitations of current user interface technology. Agents don’t have to be “intelligent” but it is generally accepted that a computer process must have the ability to be both proactive and reactive in relation to its environment, and be able to communicate with other processes in order to be classified as an agent. As we moved to developing link services for the Web, what were the information and processing filters in Microcosm evolved into systems of communicating agents.

The EU-funded MEMOIR project [De Roure et al, 2001] was the first project in which we used agents to customise the information presented to the user. In our next project, we moved to using the standard agent communication language KQML [El-Beltagy et al, 1999] in a multi-agent framework. The agents implemented as part of this project included a personal interface agent, an organisational memory agent and a link service agent. Other existing information management services, such as search engines, were integrated into the system using wrapper technology.

We built-on this architecture in the QuIC (Queries in Context) project [El-Beltagy et al, 2001]. This project built on the ideas of “linking in context” that were embodied in Microcosm. In QuIC, the associative links offered to the user when they look at a document depend on the context of the document, determined using standard text analysis techniques, and the context of the user, which in this case is the information that the system knows about the user, such as their browsing history. This project demonstrated the effectiveness of basing the architecture of such a system on a multi-agent framework.

Moreau [2000] describes the SoFAR (Southampton Framework for Agent Research) multi-agent framework which we designed and implemented to support distributed information management. It is novel because the agent communication mechanism is derived from distributed computing techniques but takes on board the reality of agency. Because of the opportunity they provide for the reuse of agent services, we hypothesise that in the longer term agent frameworks such as SoFAR will demonstrate emergent properties, as agents selectively utilise other agent services to answer particular queries. Some twelve years after the Microcosm project began, from the hypermedia perspective this agent-based approach to distributed information management represents the third generation of the ideas embodied in that project.

My current research work aims to integrate advanced knowledge technologies, including the development and use of ontologies, to enable the agents to reason, adapt and learn about the environment in which they exist, and to develop tools to support the development of the Semantic Web. This work is being taken forward under the umbrella of the EPSRC funded Advanced Knowledge Technologies (AKT) Interdisciplinary Research Collaboration which is led by Southampton, and one of the tools developed as part of this project won the 2003 Semantic Web Challenge at the ISWC in Florida, USA. An interesting project that has spun out of this work is the Arteqakt project, which is a collaboration between AKT, Equator (another EPSRC IRC that we are involved with), and an EU project called Artiste. In Arteqakt we are seeking to automatically extract knowledge from the Web, populate a knowledge base using an ontology appropriate to the subject domain, and use it to generate personalised narratives (Alani et al, 2003). There are significant pointers in this work to the development of future user interfaces. When you ‘talk’ to your personalised agent, you don’t want a list of links (a la Google) as a reply, you want a narrative that is customised to you and the context you are currently operating in.

One of the main motivations for my original work in the Microcosm project was Vannevar Bush’s paper “As We May Think” published in the Atlantic monthly in 1945. In this paper, Bush foresaw today’s global network of interconnected documents, and discusses possible methodologies for retrieving information stored in this way. He draws an analogy with the way the human brain is able to retrieve information by some sort of associative mechanism. I have always been striving to explore this idea in my research work. Can we build systems that are capable of “associative thought”? Increasingly I am interested in the way that the life sciences, in particular neuroscience and psychology, and the physical sciences can work together to enable us to better understand how the brain works and to build more intelligent artificial systems (O’Hara, 2003, O’Hara, 2004).

In some sense I feel that only now, as we move into the era of the Semantic Web and projects, such as the “Memories for Life” Grand Challenge, that have grown out of the Foresight Cognitive Systems project, are the ideas that were the driving forces behind the design of Microcosm actually being realised.

Significant publications to illustrate research track record (see full list)

Fountain, A., Hall, W., Heath, I. & Davis, H. "Microcosm: An Open Model for Hypermedia with Dynamic Linking" in Hypertext: Concepts, Systems and Applications, Proceedings of ECHT'90, Paris, November 1990, (eds. A. Rizk, N. Streitz & J. Andre) Cambridge University Press, pp 298-311 (1990).

Davis, H., Hall, W., Heath, I., Hill, G. & Wilkins, R. "Towards an Integrated Information Environment with Open Hypermedia Systems" in the Proceedings of the ACM Conference on Hypertext (ECHT'92), Milan, November 1992, ACM Press, pp 181-190 (1992).

Hill, G., Wilkins, R. & Hall, W. "Open and Reconfigurable Hypermedia Systems: A Filter Based Model" Hypermedia, Vol 5, No. 2, pp 103-118 (1993).

Davis, H.C., Knight, S. & Hall, W. "Light Hypermedia Link Services: A Study of Third Party Application Integration". In the Proceedings of ECHT'94, ACM Press, pp 41 - 50 (1994).

Hall, W. “Ending the Tyranny of the Button” IEEE Multimedia, Vol.1, No.1 pp 60-68 (1994)

Woolf, B.P. & Hall, W. "Multimedia Pedagogues: Interactive Systems for Teaching and Learning". IEEE Computer, Vol.28, No.5 pp 74 - 80 (1995).

Carr, L.A., De Roure, D.C., Hall, W. & Hill, G.J. "The Distributed Link Service: A Tool for Publishers, Authors and Readers". Proceedings of the Fourth International World Wide Web Conference: The Web Revolution, Boston, MA, December 1995.

Lewis, P.H., Davis, H.C., Griffiths, S.R., Hall, W. and Wilkins, R.J. "Media-based Navigation with Generic Links. In Proceedings of the Seventh ACM Conference on Hypertext, ACM, March 1996, pp215-233.

De Roure, D.C., Hall, W., Davis, H.C. and Dale, J. "Agents for Distributed Multimedia Information Management'. In Proceedings of PAAM'96, London, March 1996.

Goose, S., Dale, J., Hill, G. J., De Roure, D. and Hall, W. "An Open Framework for Integrating Widely Distributed Hypermedia Resources". In Proceedings of the IEEE International Conference on Multimedia Computing and Systems, Hiroshima, Japan, June 1996, pp 364-371

De Roure, D.C., Carr, L.A., Hall, W. and Hill G.J. A Distributed Hypermedia Link Service. In Proceedings of the Third International Workshop on Services in Distributed and Networked Environments (SDNE96), Macau, June 3-4, 1996, IEEE Computer Society Press, 1996, pp156-161.

Hall, W., Davis, H.C. and Hutchings, G.A. "Rethinking Hypermedia: the Microcosm Approach". Boston USA, Kluwer Academic Press, 1996, 195pp.

Hitchcock, S., Carr, L., Harris, S., Hey, J. & Hall, W. "Citation linking: improving access to online journals". Proceedings of Second ACM conference on Digital Libraries, Philadelphia, pp115-122, 1997

DeRoure, D. & Hall, W. "Distributed Multimedia Information Systems". IEEE Multimedia, Vol. 4, No. 4, pp 68 - 73, 1997

Carr, L., DeRoure, D., Davis, H., and Hall, W. “Implementing an Open Link Service for the World Wide Web” World Wide Web Journal, 1(2), 1998, 61-71.

Carr,L., Hall, W., and Hitchcock, S. “Link Services or Agent Services?” Proc. of the Ninth ACM Conference on Hypertext, Pittsburgh, 1998, 113-122.

Lowe, D. and Hall W. (1999) “Hypermedia and the Web: an Engineering Approach.” Wiley

El-Beltagy, S., DeRoure, D. and Hall, W. (1999) “A Multiagent system for Navigation Assistance and Information Finding.” Proc The Fourth International Conference on the Practical Application of Intelligent Agents and Multi-Agent Technology p.281-295.

Moreau, L., Gibbins, N., DeRoure, D., El-Beltagy, S. Hall, W., Hughes, G. , Joyce, D. et al. (2000) “SoFAR with DIM Agents: An Agent Framework for Distributed Information Management.” Proc The Fifth International Conference and Exhibition on The Practical Application of Intelligent Agents and Multi-Agents

Hall W. (2000). “The Button Strikes Back”. The New Review of Hypermedia and Multimedia. Volume 6. p5-17.

De Roure, D., Hall, W., Reich, S., Hill, G., Pikrakis, A. and Stairmand, M. (2001) “MEMOIR - an open distributed framework for enhanced navigation of distributed information.” Information Processing and Management, 37 p.53-74.

Carr, L, Hall, W, Bechhofer, S and Goble, C (2001). “Conceptual Linking: Ontology-based Open Hypermedia.” Proc the Tenth International World Wide Web Conference. Hong Kong, 1-5 May. P 334 –342

El-Beltagy, Samhaa, Hall, Wendy, Roure, David De and Carr, Leslie (2001) Linking in Context . In Proceedings The Twelfth ACM Conference on Hypertext and Hypermedia (Hypertext '01), p 151-160.

Alani, H., Kim, S., Millard, D.E., Weal, M., Hall,W., Lewis, P.H. and Shadbolt, N.R. (2003) “Automatic Ontology-Based Knowledge Extraction from Web Documents”. IEEE Intelligent Systems 18(1), p 14-21.

O’Hara,K., Hall W., van Rijsbergen K. and Shadbolt, N. (2003) “Memory, Reasoning and Learning” Foresight Cognitive Systems Project Research Review ( Now available in “Cognitive Systems – Information Processing meets Brain Science” edited by R. Morris, L. Tarassenko and M. Kenward. Elsevier (2005)

O’Hara K., Morris R., Shadbolt N., Hitch, G. and Hall W. (2004) “Memories for Life”. In preparation for Interface, the new Royal Society journal for interdisciplinary research

Wendy Hall is interviewed by Stephen Ibaraki distinguished expert and currently President of the Canadian Information Processing Society, for his regular CIPS interview series

MSc in Artificial Intelligence numbers double in course’s tenth year

Our MSc in Artificial Intelligence attracts a surge of interest a decade on from its launch

Time is right to capitalise on artificial intelligence

Just weeks after the publication of the UK government's review of the country's capabilities in artificial intelligence, Professor Dame Wendy Hall says time is right to capitalise

Dame Wendy Hall to lead major review of Artificial Intelligence in the UK

Professor Dame Wendy Hall is to lead a major UK government review on how to capitalise on the country’s status as a world-leader in the science underpinning Artificial Intelligence (AI) technology.



(c) 2009 Wendy Hall | Contact me