Conflict Watcher

Master thesis project focused on developing an open-source tool, referred to as Conflict Watcher, focused on detecting, monitoring and analysing edit wars within Wikipedia articles, with the aim of extracting information that can be correlated with real-world geopolitical conflicts. Said tool implements the mutual reverts-based detection method presented by Sumi et al. in Edit wars in Wikipedia to detect the presence of edit wars.

Features

• Search of Wikipedia articles by keywords and related articles
• Edit wars detection through the mutual reverts-based method presented by Sumi et al.
• Temporal representation of edit wars evolution
• Extraction of conflict-related information (severity, size, top mutual reverts, top reverted revisions...)
• In-depth analysis of users and revisions of selected articles
• SQLite storage and sharing of results
• Fine-grained data analysis through SQLite queries
• Automatic (and configurable) monitoring of edit wars within articles, generating a system alert when detected
• Support for Windows and, theoretically, Unix, although the latter has not been tested
• Compatible with articles from any language supported by Wikipedia
• Real-time adaptability to most screen resolutions

Requirements

• Python 3.13+ (including SQLite)
• Packages listed in requirements.txt

Provided build execution

1. Clone this repository:

   git clone https://github.com/CyberSpace-Architect/TFM.git
   

2. Open build/exe.*** directory (*** must be changed for the name of the directory within)

   cd build/exe.*** 
   

3. Execute Conflict Watcher.exe (or double-click it):

   ."Conflict Watcher.exe"
   

For developers >>> Build own version after changes

1. Activate provided Python virtual environment. From base project directory:

   .venv\scripts\activate
   

2. From the same directory, execute your own version of the program with:

   python -m app.main
   

3. To create your own build, execute (with Python virtual environment activated):

   python setup.py build