AI Powered Criminal Network Analysis System
Criminal investigations increasingly involve large volumes of linked data such as call records, financial transactions, case reports and social media activity, yet investigators still analyse these relationships largely by hand. This paper presents the design of an AI powered criminal network analysis system built in Python. The system ingests structured and unstructured records, extracts people, organisations, phones, locations and accounts using natural language processing (spaCy), builds a relationship graph (NetworkX), measures the importance of each actor with centrality metrics, detects groups with community detection, suggests hidden or missing links with machine learning, and presents the result as an interactive graph with an investigator-readable report. The system is organised as a linear five-stage pipeline that is modular and can be run on modest hardware. The paper describes the architecture, technology stack, processing flow, applications, advantages and limitations, and discusses ethical and privacy safeguards, a planned evaluation methodology and future work including graph neural networks, multilingual (Hindi and Marathi) text processing and temporal network analysis. No measured performance results are claimed in this paper
Authors
- H. B. Jadhav, Dr. H. B. Jadhav, Dr. H. B. Jadhav
- Pravej Raju Attar Rohit Rajesh Ghallal
- Rohit Rahul Chayal Abhishek Balasaheb Gaikwad
Publication Details
- Journal
- International Journal of Advanced Research in Science Communication and Technology
- Published
- 2026-11-18
- DOI
- https://doi.org/10.48175/ijarsct-40830
- Primary Topic
- Crime, Illicit Activities, and Governance
- Type
- article
- Field-Weighted Citation Impact
- 0.00