Jane Reid's project suggestions 2003/04

mailto: jane@dcs.qmul.ac.uk

Areas of interest

My main interests lie in information retrieval (IR), especially user-centred issues, and the sub-area of structured document retrieval (SDR).  Have a look at the IR group Web page to see what we've been doing recently.  I also have an interest in distance learning.

Topics of particular interest at present are:

Projects on these topics are likely to take a "multi-disciplinary" approach, and draw on other areas / disciplines, e.g.

Requirements

Since this type of project will require learning quite a few new skills, some prior knowledge of the following areas MAY be an advantage: Contrary to some people's expectations, projects in these areas are NOT easy (ask any of my previous project students!), but they do require different skills from traditional programming-based projects.

Typically, a project like this might involve:

Specific projects

In traditional information retrieval systems, documents are retrieved as atomic units.  However, documents often display a logical structure.  For example, a large corporate report will contain a title page, introductory material and several chapters, each composed of sections (this may be stored in mark-up languages such as SGML).  Another example of highly structured documents are web documents, where the structure can be internal (HTML or XML), and external (documents connected by hyperlinks).  From the user's point of view, presenting only some components of documents can make it easier to distinguish potentially relevant documents from irrelevant ones.  It can also make it easier for a user to target components of a document that may be most useful, especially for long documents, and documents that cover a variety of subjects.  This is the aim of structured document retrieval.  A further development of structured document retrieval, called focussed retrieval, aims to support users in their information-seeking behaviour by presenting them with best entry points (BEPs), i.e. only those document components that provide the best starting-points for browsing to relevant document components.

Possible project topics in this area include the following at present. Most will require reasonable programming skills, and will involve getting to grips with XML at a basic level.  They will also involve learning some basic experimental design and statistical analysis, so you should have taken (and enjoyed!) the GUI module.
  1. Examining different types of BEPs, including user criteria for BEP selection, and the usage and effectiveness of the different types.
  2. Exploring a possible relationship between query type/complexity and BEP type.
  3. Work on BEPs in the context of new data we have recently gathered (during INEX).
  4. Exploring a possible relationship between query type/complexity and interface designs to support information seeking behaviour.
  5. Extending a specific, existing interface (the RLB interface) to accommodate browsing within an individual document, as well as browsing between documents.
  6. Interface design for SDR systems to incorporate information regarding ranking and structural information for individual document components, as well as relationships between document components.
Other possible projects concern distance learning, especially the application of information seeking and retrieval in this context. Possible topics in this area might relate to the following (but are very flexible):
  1. Design and implementation of a virtual learning environment, using open source software, to support distance learning in the Open and Distance Learning Unit.
  2. Evaluation of virtual learning environments.
  3. Multimedia information retrieval for distance learning.

Interested?  Want to find out more?

Mail me to make an appointment.

Also feel free to propose an idea of your own around the topics described above.