الملخص الإنجليزي
Campus portal is the entry point for a college or university providing access to campus's electronic resources and services through authentication process depending on users' role in campus. It was aimed in this thesis to design a plagiarism detection tool among java programming assignments and plug this tool into a campus portal through a learning management system (LMS). The tool was designed using the attribute counting metric, where ten metrics were used in the study to compare the assignments pairs and hence produce similarity ratios to measure plagiarism. The tool provided good results among the used set of data.Up on the previous aim, a study of the concept of campus portals and the ability to shift the SQU campus Website into a campus portal. Up on the study made, it was found that the SQU is not yet ready for the shift and more concentration need to be made on developing the resources and services provided by the SQU intranet. Analysis and design of a proposed SQU campus portal was made with a study of the analysis and design of the WebCT (LMS). The tool was then plugged into WebCT LMS as part of it, and then ran successfully on a set of data.The examination made on the metrics used by the tool, showed that the weight of a metric in the calculated similarity ratio is affected by the use of that metric through the program. Metric's weight increases by the increase of its use through a program. Hence, keywords and methods are the metrics with higher weights. As well, comparison was made between the tool designed by the current research and another tool (JPlag), which uses the structure metric approach. Due to that comparison, it was found that results of both tools were close in most sets of the data used in the study. The designed tool gave better results in cases of very close copies, and assignments aimed to novice programmers than to experience ones. JPlag gave better results in more experienced programs. JPlag, also better detected pairs with simple changes in naming, loop statements and use of redundant statements, than the other tool. While pairs with changes in order of lines were better detected by the research-designed tool.
