CallMine

CallMine: Fraud Detection and Visualization of Million-Scale Call Graphs

Authors: Mirela Cazzolato^1,2, Saranya Vijayakumar¹, Meng-Chieh Lee¹, Catalina Vajiac¹, Namyong Park¹, Pedro Fidalgo^3,4, Agma J. M. Traina², Christos Faloutsos¹

Affiliations: ¹ Carnegie Mellon University (CMU), ² University of São Paulo (USP), ³ Mobileum, ⁴ ISCTE-IUL

Work accepted for publication at CIKM'2023

Setup environment

To create and use a virtual environment, type:

• python -m venv wcw_venv
• source wcw_venv/bin/activate

To install the requirements:

• pip install -r requirements.txt or simply make prep

Usage:

Type make demo to see a demo of CallMine and CallMine-Focus

Sample Dataset:

File INPUT_DATA/sample_raw_data.csv' has a synthetic data sample with:

11,000 calls, consisting of

• 10,000 random calls:
  • 2,000 sources
  • 2,000 destinations
  • 5 days
• Cluster with 1,000 calls
  • 10 sources
  • 10 destinations
  • Phone calls with duration between 20 and 40 seconds