pip-db - Final Project Report

Mark attained: 90%.

Formal feedback

This is a truly exceptional piece of work of extremely high quality, insight and diligence. An appropriate lifecycle model has been followed in depth; new development tools created (that could have formed a project in their own right); a highly user-centred design approach was followed, leading to strong user approval of the product; a full prototype was implemented; the student learned a new programming language (Clojure) and mastered it in a few weeks. To give a measure of the scale of the achievement, 2,420 code revisions were committed and over 34,000 lines of code written. The only real criticism would be that all the toolwork and change of programming language restricted the time available for development of functionality: the lack of tools to add to the dataset is a loss. However, given the scale of achievement, this is a minor quibble. A lot of deep thought has gone into every decision made during this project, and this is clearly conveyed in the report. Quantitative and qualitative measures are used to assess the impact of decisions (for example, a reduction of 75% in the size of the code base when moving from PHP to Clojure). The use trials were a bit limited (just 5 users) as the student recognized himself. The report is also a great piece of work. Exceptionally clearly written and grammatically sound (I counted fewer than 5 errors in 67 pages), it makes good use of diagrams and really explains some difficult concepts with excellent examples and code fragments. All the practical work is backed up by theory drawn from the literature (there were 51 references). This is a very well written report on an outstanding project with exceptional work approximating a professional standard and contribution to the knowledge of database design for medical applications. The end-product is characterised by a very high standard of functionality and usability. A novel application of Markov chains has been used in generating the test data. The end-product has been thoroughly reviewed and critically analysed based on appropriate theoretical references, however it would be better to involve the comparison with other similar products in the evaluation. The standard of writing is close to professional standard, with nearly no flaws in English, typography (just one Chapter number missing on page 2) or presentation. A lot of relevant academic references are used. Diagrams and tables are appropriately used for illustration. Just a couple of suggestions: the Chapter titles can be a bit more detailed showing what process, infrastructure and evaluation etc. The caption of some figures (Figure 4.2, 5.4 – 5.10) can be more concise, with detailed description/explanation in the main text instead of in the figure caption.

A full time geek and research student with a passion for developing great software, often late at night.