The Deeside Insurance WBA Project



The Deeside Insurance Web Based Application (WBA) project is a Knowledge Transfer Partnership (KTP) between:

The Department of Computer Science at the University of Liverpool, and
Deeside Insurance Limited.

The project is funded under the DTI Knowledge Transfer Partnership scheme. The overall objective of the project is to develop, and embed in the company, a robust tool kit for building Web Based Applications and to launch a series of insurance web sites and affinity schemes. A two year programme of work is envisaged comprising seven stages each as follows:

  1. Background and Reviewe: The work will commence with necessary background work to allow familiarization with insurance industry business processes/current IT practices. Deeside Insurance uses software systems supplied by OpenGI Ltd to access current quote information supplied by most major insurance companies. This database can be accessed using an appropriate software interchange format (written in XML). The associate is also expected to review the current state of the art in Web Base Application (WBA) development (including a literature search of relevant research), considering current tools and techniques relevant to the building of e-commerce web sites relevant to this project. All ideas will be assessed for relevance and potential for adaptation. The current IT infrastructure of the company will also be reviewed by the Associate in consultation with senior management, to establish the IT requirements to deliver project and future outcomes and impacts.
  2. Search Engine Optimization: Review current thinking on SEO and produce an SEO strategy document to be referenced when constructing the initial WBAs (for company and affiliate customers) and, also, the WBA tool kit. Review and update the SEO strategy document in the middle and at the end of the project, briefing Management on any key changes.
  3. Building and Launch of Initial WBA: The WBA will be for car insurance. The reasons for this choice are: (i) car insurance represents the largest market segment for the company; (ii) the company already has a legacy car insurance WBA (which lacks functionality); (iii) car insurance is likely to be the segment most in demand for affinity schemes.
  4. Development of a WBA for a Key Affinity Customer: The Associate will develop and set up an affinity scheme (white labelling) insurance web site for an immediate customer (e.g. a national retailer or a premier league football club) using, as the basis, the WBA developed in tasks 9 and 10. A key objective will be to identify necessary additional functionality/requirements for the affinity scheme that will need to be reflected in the capability and versatility of the toolkit (stage 6). The Associate will engage with the affinity scheme customer and staff through to the launch of the WBA.
  5. Web Usage Mining: Customer web usage will be investigated by the Associate using data mining technology. The focus will be on the way that users engage with: (i) the existing (legacy) WBA for car insurance; (ii) the WBA developed in task 9; (iii) the developed affinity system. Established Web Mining techniques to cluster/classify users of the WBAs. This stage will commence with pre-processing of web log data to which existing clustering and classification algorithms will be applied. The aim will be for the Associate to identify and interpret usage patterns which can be usefully applied to influence organisation of WBAs.
  6. Development of a WBA Tool Kit: This is the main objective of the work. The development process will comprise: (i) Tool Kit Requirements, (ii) Tool Kit Specification and Design, (iii) Tool Kit Implementation and (iv) Tool Kit Evaluation. Requirements will be founded on on experience obtained producing the car insurance WBA, the affinity scheme application and the web mining results. The toolkit/template editor will probably take the form of a Graphical User Interface (GUI) supported by a sophisticated database. The Associate will produce design diagrams and story boards illustrating operational views of the system. The insurance toolkit and template editor will be implemented using an appropriate development strategy coupled with module and system testing. The most reliable development strategy is likely to be an Object Oriented (OO). Evaluation will be undertaken by producing and launching a 'suite' of 4 insurance WBAs including 2-affinity schemes. The evaluation will be conducted conjointly by the Associate, the knowledge base partner and members of staff at the company.
  7. Staff Training and Future Development: The insurance WBA tool kit created within the project will continue to be maintained/developed on completion of the KTP project by in-house staff at Deeside Insurance. This will be enabled by the Associate through: (i) the production of a user manual and (ii) staff training. Through dialogue with management, staff and clients, the Associate is also expected to explore functionality appropriate for the tool kit to support business growth.


The project team currently comprises (May 2008):

  1. Shane Williams (company supervisor and LMC chair) - Deeside Insurance Ltd.
  2. Frans Coenen (lead academic and academic supervisor) - Department of Computer Science, the University of Liverpool.
  3. Tim Williams (company facilitator) - Deeside Insurance Ltd.
  4. Katie Atkinson (academic support) - Department of Computer Science, the University of Liverpool.
  5. TBA (KTP associate) - Department of Computer Science, the University of Liverpool.

The KTP advisor is Idris Price.


(Local access only.)


Job Advert.

Created and maintained by Frans Coenen. Last updated 1 May 2008.