An empirical evaluation of modelling languages for fourth generation e-commerce application development
The rapid development of Web technologies and with it the increased demand for
quick Web application development has taken the Web development process into a
new state. Several academic papers and surveys describe the current state of the Web
market as similar to the `software crisis' of the early 60s, when software development
practices were ad hoc and unstructured. Developers are realising the need for control,
management, and sound modelling languages to document Web applications.
This research investigates and empirically evaluates ten of the main Web modelling
languages available to the market. A set of eight criteria is identified from the
literature. These criteria are then fine tuned and verified in a questionnaire survey
targeting fourth generation e-commerce companies in the UK. Respondents of the
survey suggest a ninth criterion, which is further verified and added to the list. The
author implements these Web models using a real case study, the IPCIS Web
An empirical evaluation process takes place whereby the outcomes of the Web
implementation are evaluated against the verified criteria. The evaluation process
helps to verify the research hypothesis and establishes that none of these models
meets the market requirements. The recommendations of the evaluation process
provide insights into what is missing in these Web models.
The survey outcomes are tested for sampling error and statistically projected onto the
population as a whole. From these projections, this research establishes that the
majority in the market have heard of at least one of the Web models; however, the
majority do not use them. Finally this research concludes that the market does believe
in the need for Web models yet are disappointed with the current models on offer.