Muhammad Shuja Undergraduate Dissertation 2017/18

JavaGrader: An online system for automatic marking of Java assignments

Supervised by S.North

Abstract

In a world which changes everyday, technology plays a vital role in keeping it spinning. The lack of competent and creative programmers does not really help the technology sector when it comes to solving challenges like client privacy, Big Data and Artificial Intelligence. In order to gain competence in the skill, the right foundations need to be set. This leads to the need of academic institutions improving their learning and teaching strategies. Evaluation techniques need to be catered more towards a student developing their skills through practice and feedback.

This report discusses the inspiration, aims and objectives and design of the JavaGrader: an online client-server based architecture for automatically marking Java assignments as part of a University course. The design planning reflects on how this solutions could potentially help lecturers reduce their workload through automation, and students getting customised feedback after every submission. The implementation, however, can not be considered to be a huge success due to reasons which will be discussed in the report. Regardless, this project hopes to have still made a start towards having a better and bigger platform built through this project because of potentially having vast prospects for upgrades.