About
George Karabatis is a Professor of Information Systems at UMBC. Prior to joining UMBC he was a Research Scientist at Telcordia Technologies (formerly Bellcore) working on information related research for the telecom industry.
Research interests
Semantic data science, semantic information integration, semantics for AI, machine learning, heterogeneous data integration, cyber-security, intrusion detection and prevention, databases, entrepreneurship.
Teaching interests
Data science, information integration, databases, and cyber-security courses.
Education
- Ph D, Computer science — University of Houston (1995)
- MS, Computer science — University of Houston (1988)
- BS, Mathematics — Aristotle University of Thessaloniki (1983)
Publications
- Adversary-Resilient Clustered Federated Learning for Secure AI-Driven Healthcare Data Analytics 2025
- On the Verification of Software Vulnerabilities During Static Code Analysis Using Data Mining Techniques 2017
- Queryable Semantics to Detect Cyber-Attacks: A Flow-Based Detection Approach 2016
- Context Aware Discovery in Web Data through Anomaly Detection 2015
- Context and Semantics for Detection of Cyber-Attacks 2014
Grants and Contracts
- Anonymization of Network Trace Data through Differential Privacy 2016
- Energy Education through Green Buildings 2015
- Sensor Technology Box for Smart Health 2015
- Anonymization of Network Trace Data 2015
- TWC: TTP Option: Small: A Flow-based Detection of Cyber-attacks using Context in Semantic Link Networks 2015
- Semantic, contextual, and scalable detection of zero-day attacks for cloud environments 2015
- The Automated cyber-attack PredictiOn, deteCtion and AnaLYtic Processing SystEm (APOCALYPSE) 2015
- Energy Education through Green Buildings 2014
- TWC: Small: Leveraging Context in Semantic Link Networks to Detect Cyber-attacks: A Flow-based Detection Approach 2014
- CRAFT: Contextually Revealing Attacks over Flows Tool 2014
- Energy Education through Green Buildings 2013
Patents and Intellectual Property
- A method to generate mappings of heterogeneous relational schemas using unsupervised learning 2022
- A METHOD FOR ANONYMIZING NETWORK DATA USING DIFFERENTIAL PRIVACY 2022
- A MACHINE LEARNING AND DATA SCIENCE METHOD FOR IDENTIFYING TRUE AND FALSE POSITIVE VULNERABILITIES IN SOFTWARE SOURCE CODE 2019
- FLOW-BASED SYSTEM AND METHOD FOR DETECTING CYBER-ATTACKS UTILIZING CONTEXTUAL INFORMATION 2014