ABET Program Certification
HCI 551 Project 2

download Project Documentation (.pdf)

This evaluation project was a very ambitious undertaking. Not only did it entail the implementation of two vastly different prototypes using related, but incompatible technologies, it also entailed the evaluation of these prototypes and – primarily – the interaction scheme of information items the prototypes are intended to visualize. Hence, this project was essentially two projects in one: The implementation portion and the evaluation.

The implementation turned out to be significantly more challenging than initially expected. Since the submitting authors decided to implement the underlying database schema as well as the front end prototypes of the database, a considerable amount of time had to be invested into the aspect of the database, even though it was not immediately the subject of evaluation. But since creating a mock-up of a database schema using the information items and information interaction paradigms (whose evaluation was discussed in the earlier section) was just as difficult as implementing it fully, and hence delivering a deployable product was possible as an extra benefit for the project stakeholders, the submitting authors committed to completeness rather than theoretical mock-ups.

The implementation stage of this project was also rather challenging, as technologies were selected by the authors that were not fully established to deliver the goal that was intended by the project stakeholders. While Java based technologies have in the past proven to be well-qualified to be used with a database back end for form-like information presentation, especially applets and the Google Web Toolkit reserve challenges that needed to be addressed. For instance, as both technologies are made available through the browser, security policies regarding applets disallow basic I/O operations and network connections, making it difficult to connect to remote database servers. Furthermore, this approach requires the use of database drivers, which must be available to the client without the need of setting up program dependencies or class path variables. The biggest challenge for GWT lied in the interfacing with a database as connectivity can only be achieved using asynchronous Remote Procedure Calls. Setting these up is a major undertaking when deploying the GWT application to a web server, which in turn must also be configured accordingly. Another issue was the premature state of the GWT Java API. Even though GWT is currently in version number 1.5, there are many pitfalls in the API that had to be addressed by the authors in order to still achieve the intended goal.

We believe, however, that the selection of technologies was ideal for the outcome of this project. We were able to deliver not one but two fully working, deployed and customizable prototypes that address the needs and requirements that we had established in Part II of this document. The evaluation of prototype and information interaction in Phase III shows that the gathered requirements in Phase II have accurately met the needs of the three user groups and resulted in the creation of well designed prototypes. Despite the obvious flaws in the selected technologies, the choice for these technologies was a fortunate one. Since the prototypes, especially the applet approach, achieved such high regards during the evaluation, it can be safely said that these technologies are adequate to implement the ABET certification review interface that was the focus of this project. Additionally, with this project, we have established how to address the issues of the technologies, as mentioned above. This has a benefit beyond the scope of this project, as it can serve as a guideline for future projects with similar goals. We gladly share our insights and solutions to the problems, pitfalls and challenges that we have surmounted in this project to achieve excellence. We provide our insights in Appendices B through D in the form of tutorials that address the most important issues that we have encountered. Our hope is that these tutorials will help future students and researchers to overcome problems faster so that they can benefit from the work that we have done too.

To conclude, this project was extremely successful. The authors learned a great deal in two software engineering related aspects, namely software development and product testing. We learned many important lectures in terms of requirements engineering and product evaluation as well as project time management. To wrap up the project, we have selected the more successful prototype – the Java applet – for improvement and added features that were suggested during the evaluation, like a login menu and user authentication. Furthermore, we enhanced the quality of descriptions and context-sensitive help by adding Tool Tips and modified the labeling of some control elements – all in accordance with the results of the evaluation. Our endeavor was to present a complete, ready for use system.

The database and software that has been implemented throughout the project is fully documented. The authors formally make all documentation, source code, tutorials, and implementation notes available to the project stakeholders, namely the faculty of the Computer Science Department at SUNY Oswego. It is our hope that our project software will be helpful in their work and be used frequently.

Naturally, the work doesn't stop here. Even though this project is coming to an end, future work would entail a more thorough analysis of the technological nuances of the Java applet prototype versus the GWT prototype. In this project, both prototypes use different strategies to visualize the data. Evaluation has found the strategy of the applet approach to be the more successful. However, it would be interesting to see how the strategy of the applet would perform when implemented in GWT. This question could be addressed in a future project. Future work should also address a “fine tuning” of the Java applet according to exact specifications of ABET. Since these have been rather vague, they have not been taken into account in this project. Close work with ABET could help address this question.