CodeGrade’s code plagiarism detection feature can be found on the assignment management page. Here, new plagiarism runs can be done and previous runs can be examined for plagiarism.
Individual plagiarism cases should always be manually reviewed.
Running the Plagiarism Checker¶
As of writing, CodeGrade offers plagiarism detection using the JPlag code plagiarism detection tool. Alternative plagiarism tools will be supported in the near future.
Multiple settings can be set when checking for plagiarism:
- The programming language has to be selected from a list of supported languages. If the required language is not present in the list, it is no problem to select a language with similar syntax and structure.
- Suffixes, separated by commas, of the files to include can be given. Set to
.*to include files with all suffixes and leave empty to automatically select suffixes based on the chosen programming language.
- The minimal average similarity of a case to include in the results can be set. For JPlag this is a percentage:
100means exactly the same and
50means an average of 50%. This is default set to
- Finally an old assignment can be selected to include in the plagiarism check. This can for instance be the same assignment from a previous year or course. Submissions from previous assignments are only used to check with current submissions and cases between two previous submissions are not included in the results.
Additionally, instead of selecting an old CodeGrade assignment to check against, an archive with old submissions to check against can be uploaded. This should be an archive whose top level entries are treated as a separate submission. The top level entry can either be a directory or archive whose contents count as a submission, or a single regular file which counts as the entire submission.
Finally, an archive with base code can be uploaded. Code in this archive will be excluded in the plagiarism check to reduce the amount of false positives.
It is often a good practice to upload template code or provided code snippets as base code to be excluded.
JPlag is an open source software plagiarism detection tool that is developed by the Karlsruhe Institute of Technology, who describe JPlag as:
JPlag is a system that finds similarities among multiple sets of source code files. This way it can detect software plagiarism. JPlag does not merely compare bytes of text, but is aware of programming language syntax and program structure and hence is robust against many kinds of attempts to disguise similarities between plagiarized files. JPlag currently supports Java, C#, C, C++, Scheme and natural language text.
JPlag is typically used to detect and thus discourage the unallowed copying of student exercise programs in programming education. But in principle it can also be used to detect stolen software parts among large amounts of source text or modules that have been duplicated (and only slightly modified). JPlag has already played a part in several intellectual property cases where it has been successfully used by expert witnesses.
Taken from JPlag, please consult for more information.
After the plagiarism checker is done running, its results can be found under the Previous Runs tab. The run can be deleted by pressing the button or reviewed by clicking on the entry itself to find an overview page with all the potential plagiarism cases.
This overview page displays potential plagiarism cases between students, with the maximum score and the average score. The cases can be filtered on student name by using the filter bar on top.
More details can be found by clicking on the individual cases. The individual plagiarism comparison page displays the specific correspondences found between the submissions of the students. These individual correspondences are coloured differently, clicking a correspondence will display the correspondence in the code. This page can be used to review the individual plagiarism cases.
Use the plagiarism button in the sidebar to quickly toggle between plagiarism cases when reviewing individual cases.
Exporting Plagiarism Cases¶
A report for an individual potential plagiarism case can be exported from the
individual plagiarism comparison page. A
.tex file that includes the
code segments and can be supplemented with additional comments can be exported.
Select the plagiarism correspondences to be included in the report and press the Export button to generate the .tex file.
Optionally, select the Each listing on a separate page option under Options to have each listing (i.e. printed code segment) on a new page.
Make sure pop-ups from CodeGrade are allowed in your browser if downloading fails.