Problem statement

While the Internet is growing, the informational overload for individual users is increasing. People seem to increasingly struggle with the first step – making a choice of tools and information sources to integrate into their everyday workflow. For the majority of Internet users, the required learning curve and the cost of change are just too high.

Is it possible to develop a self-structuring Web – organised through users and “some smart technology”? This is the antithesis to the current and the Semantic Web: Get rid of all visual and hierarchical constraints that impose the author’s mental model onto the user. Free the users’ minds and give them the desired tools to scale, shape and assimilate information according to their needs.

Literature reviewed

World Wide Web: information, content, context, navigation and search
Visual Interfaces: various visual focus+context visualisations
User: human perception, cognition and memory, user as author (blogging)
Personalisation: Constructivism as reasoning, various techniques of personalisation, awareness and trust in personalisation
Tagging: tag-based visualisations, structured and hierarchical tagging
Briefly related to this project: Semantics, Linguistics, User modelling, Data mining, Security, Privacy

Conclusions from the literature

Current structures of the Web will not help to solve informational overload. Navigation as a concept to support cognitive processes is undervalued. Search is not the answer to our problems with the Web. Web interfaces should increasingly adapt focus+context visualisations to minimize cognitive load, improve navigation and task performance. Personalising this context “in the right way” is critical and requires and understanding of human perception, cognition and memory. Current personalisation techniques only target personalisation of content, but ignore largely personalisation of structures. This is largely due to the current architectural model of the Web. A novel tag-based architecture might allow the creation of a truly personalisable Web.

Proposed contribution

This project aims to make empirical, theoretical and methodological contributions to the field. By investigating feasibility and scalability of socially constructed personalised content interfaces, this project hopes to contribute a few steps on the way towards a novel kind of Web. One that adapts to the user, that is scalable for the user and that provides personalised contextual content. Moreover, this novel Web would simplify the authoring process by making the construction and maintenance of large scale, complex information structures as easy as publishing a blog entry. To summarize all these aspects, this project will create a generalisable theoretical framework for modular and personalised content spaces.
Furthermore, this interdisciplinary project will provide a window of opportunity to point out how to bring together different research communities (machine learning, education and HCI) to work together and achieve contributions which neither of them could achieve on their own.

What is the scope of the thesis?

The scope of this research project is the evolutionary development, testing and evaluation of prototypical content spaces for the Web. This involves the creation of a novel tag-based information architecture and various mechanisms to personalise the interface, which will apply focus+context visualisations based on the Focus-Metaphor. This will involve in depth investigations of interface usability, user experience and interaction techniques. However, this project does not aim to make substantial contributions in the field in machine learning, rather applying or combining existing approaches to a reasonable level and following a human-centred perspective.

How will the contribution be validated?

Research contributions will be validated through comparative eye-tracking experiments and longitudinal online field studies. In general, quantitative evaluations will back up investigations in relation to task performance and qualitative evaluations will back up investigations into usability and user experience. Results will be published at peer reviewed conferences and journals to ensure validity of the findings.

Want to know more? beta

Links:

- Human Centred Systems @ UCL

- SONATA project page

Publications:

Sven Laqua, Nnamdi Ogbechie and M. Angela Sasse (2007) Contextualizing the Blogosphere: A Comparison of Traditional and Novel User Interfaces for the Web. To be presented at HCI 2007, 3-7 Sept. 2007, Lancaster, UK.

Sven Laqua, Shane Udaraka Bandara and M. Angela Sasse (2007) GazeSpace: Eye Gaze Controlled Content Spaces. To be presented at HCI 2007, 3-7 Sept. 2007, Lancaster, UK.

Sven Laqua (2006) Social Construction of Personalisable Content Interfaces. Presented at Human-Centred Technology Postgraduate Workshop, 11-12 September, Brighton, UK.

Sven Laqua, Gemini Patel and Angela M. Sasse (2006) Personalised Focus-Metaphor Interfaces: An Eye Tracking Study on User Confusion. In Proceedings of Mensch und Computer 2006 , 3-6 September, Gelsenkirchen, Germany. (German Chapter of ACM)

Sven Laqua and Paul Brna (2005) The Focus-Metaphor Approach: A Novel Concept for the Design of Adaptive and User-Centric Interfaces. In Proceedings of Interact'05, 12-16 Sep, Rome
[pdf] [slides of talk] [publisher]