Real-time Anomaly Detection inHigh-Speed Time-evolving Graphs

Abstract

Real time anomaly detection in graph streams is an important task with high impact applications including cyber security, terrorist detection, fake user detection, etc. In this work, we propose to develop methods to detect anomalies in real time graph streams. For the purpose, we develop local triangle counting algorithms for graph streams, and devise anomaly detection methods based on local triangle counting. Our method has three advantages: 1) it is memory-efficient so that it can run in a system with limited memory, 2) it gives real time response to a query, and 3) it provides high accuracy. The developed method will be a crucial module for real time anomaly detection in high speed time evolving graphs.

Document Details

Document Type
DoD Grant Award
Publication Date
Sep 21, 2018
Source ID
FA23861614044

Entities

People

  • U. Kang

Organizations

  • Air Force Office of Scientific Research
  • Seoul National University
  • United States Air Force

Tags

Fields of Study

  • Computer science

Readers

  • Distributed Systems and Data Platform Development
  • Graph Algorithms and Convex Optimization.
  • Neural Network Machine Learning.

Technology Areas

  • Cyber
  • Cyber - Cryptography