JISC

Summary

E-Research can be defined as the use of networked, distributed and shared digital tools and data to support the research process for the production of knowledge. It is designed as both a conceptual guide to the creation of an “operating philosophy” for research processes using the web and as a practical guide for educational researchers. A major drawback of e-research tools is their inability to provide a variety of simple and complex interfaces and services depending on whether the researcher is experienced or novice researcher[1]. The e-Research adaptive user interface (eRaUI) project is aiming at developing a personalized user interface for a NaCTeM text mining e-Research tool (http://www.nactem.ac.uk/). eRaUI will be adaptable to different usages and different level of researchers’ knowledge and preferences increasing the use of NaCTeM e-research tools by making it easier to learn and adaptable to the requirements of different user groups.

To build an adaptable and learnable eRaUI we will combine different algorithms and techniques. We will use inductive algorithms from machine learning to build knowledge bases (user models) based on the observation of the users’ activities. This captured knowledge will be used to make the interface more informative and the users searches more effective. We will implement content and collaborative based filtering methods, as the basis for the user to select and learn about digital content of NaCTeM. We will enhance eRaUI interface with workflow systems capabilities to model NaCTeM research processes and tasks. These workflow features, combined with the user model interface, will enhance eRaUI capabilities to generate actions and plans (generative interfaces) to guide the users in their learning process and guide them through their research process, exposing appropriate functionality as needed i.e. providing an adaptive minimal interface or “onion” models of user interfaces in which each layer exposes slightly more complexity providing a graduated learning curve to novice researchers. We will also incorporate visualisation techniques to help users develop and widen their search by transforming the rich textual interface of the NaCTeM search engine into an information visualisation where search results are displayed in window panes or an interlinked network.

Objectives

The main aim of this proposal is to develop an adaptive and learnable e-Research User Interface (eRaUI) for NaCTeM e-Research tool. This aim will be achieved through the following objectives: (i) identify different types of users of NaCTeM and carry out user modelling, (ii) implement advanced filtering algorithm(s) for extracting appropriate digital content to match user needs, (iii) analyse and design adaptive UI with workflow system capabilities to support different groups’ activities and learning (iv) use visualisation techniques for better navigation, information display and action planning and (v) integrate eRaUI with NaCTeM, implement, test and evaluate its adaptability and learnability.

Anticipated Outputs and Outcomes

The project will produce:

• An adaptable and learnable interface to NaCTeM text mining e-research tool called eRaUI that significantly improves the learnability of NaCTeM resources;
• A project blog that documents the progress of the project with at least monthly postings;
• A detailed briefing paper that describes the approach taken and the lessons learned in terms of learnability and usability design;
• A final budget and completion report.

[1] Langley, P. (1997). Machine learning for adaptive user interfaces. Proceedings of the 21st German Annual Conference on Artificial Intelligence (pp. 53-62). Freiburg, Germany: Springer.