03/27/2026
By Manasvi Malhar Sudershan

The Miner School of Computer and Information Sciences, Department of Computer Science, invites you to attend a master's thesis defense by Manasvi Malhar Sudershan on "Mobishield: Detecting Dual Use Surveillance Apps on Android Devices."

Date: April 10, 2026
Time: 10:15 a.m.
Location: Wannalancit Mills 445 and via Zoom

Committee Members:

  • Advisor: Sashank Narain, Ph.D., Assistant Professor, Miner School of Computer and Information Sciences
  • Ian Chen, Ph.D., Assistant Professor, Miner School of Computer and Information Sciences
  • April Pattavina, Ph.D., Professor and Chair, School of Criminology and Justice Studies
  • Claire Lee, Ph.D., Associate Professor, School of Criminology and Justice Studies

Abstract:
Technology facilitated abuse is one of the fastest growing forms of intimate partner violence. Mobile devices store highly sensitive personal information, making them a prime target for surveillance and privacy invasive applications. Stalkerware and hidden monitoring applications exploit system permissions and device level access to track user activity without informed consent. Detecting such applications remains challenging due to obfuscation techniques, dual purpose functionality, and evolving malicious behaviors.

This thesis presents the design and implementation of an Android based anti stalkerware detection system that identifies hidden and potentially malicious applications through permission analysis and cloud based threat intelligence. The application performs comprehensive scanning of all installed apps, extracts their permission sets, and analyzes behavioral indicators. To enhance detection reliability, a backend Flask server integrates with the VirusTotal API to perform multi engine malware evaluation. The results are processed and returned to the mobile device, allowing users to view risk assessments in a structured and interpretable format.

The system integrates a Large Language Model (LLM) based chatbot that enhances the interpretability of security analysis results. Unlike traditional security tools that present raw and often complex technical outputs, the chatbot provides context-aware explanations of application behavior, permission usage, and identified risk factors. By leveraging natural language processing capabilities, it translates technical findings into clear, user-understandable insights and actionable recommendations. This not only improves accessibility for non-technical users but also empowers them to make informed decisions regarding application safety and device security.

The proposed solution combines local application analysis, external threat intelligence, and conversational AI support to enhance mobile security transparency. The system demonstrates a practical, deployable approach for strengthening user awareness and identifying surveillance related threats in Android environments.

All are welcome to attend.