Misconduct Assessment Tool (MAT)

Tool for accessing academic miscoduct

The aim of this tool is to test student submissions that are not obvious cases, but suspected of misconduct due to multiple small segments being similar to other submissions. With a large number of submissions, it is fairly likely that some segments will be similar by chance. This tool is used to help in such situations, by estimating the expected number of submissions with that exact combination of segments. It is built to assist the decision making process, it does not replace it.

For each suspect segment, the individual probability of that segment is calculated by taking the number of submissions that have a similar segment over the number of all the submissions. Segments that are present in a lot of submissions have a higher individual probability. The joint probability is then calculated by assuming that each segment is independent and taking the product of all the segments' individual probabilities. Finally, to calculate the estimated number of submissions we take the joint probability and multiply it with the number of all the submissions. The resulting number should be the number of students anticipated to have the exact combination of the suspected segments. If the expected number is lower than one, a case of misconduct is suggested. Please note that the results are not conclusive.

Requirements

Java SE Runtime Environment(JRE) >= 1.8.0

Optional

Anaconda >= 5.2 OR Miniconda >= 4.5.4

Otherwise, install the required Python dependencies listed in environment.yml using pip manually and skip the the installation guide below

Development Installation

1. Clone the repository

git clone https://github.com/iamstelios/Misconduct-Assessment-Tool

2. Create the conda environment:

cd Misconduct-Assessment-Tool/scripts/
conda env create -f environment.yml
cd ..

3. Activate the environment

OS X & Linux:

conda activate mat

Windows:

activate mat

Note that the system requires an environment variable SECRET_KEY to be set

4. Follow the link below for instruction of how to include the SECRET_KEY environment variable in your conda environment. https://docs.conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html#saving-environment-variables

Running the Misconduct Assessment Tool

1. Apply the database migrations (only when the database model structure is modified and during the first run)

python manage.py migrate

2. Run the server!

python manage.py runserver <optional ip:port>

Screens

Welcome Page

Upload Page

Segment Selection Page

Results Page

Meta

Stelios Milisavljevic – https://iamstelios.com – https://github.com/iamstelios

Distributed under the GNU General Public License. See LICENSE for more information.

Final year project UG4 BEng Software Engineering

Supervised by Prof. Kyriakos Kalorkoti, School of Informatics, University of Edinburgh

System inherited from Yuechen Xie – https://github.com/Weak-Chicken/misconduct_detection_project

Contributing

1. Fork it (https://github.com/iamstelios/Misconduct-Assessment-Tool)
2. Create your feature branch (git checkout -b feature/fooBar)
3. Commit your changes (git commit -am 'Add some fooBar')
4. Push to the branch (git push origin feature/fooBar)
5. Create a new Pull Request