Territorial control and violent non-state actors
Recent work examines how territorial control transforms the capabilities and vulnerabilities of violent organizations.
University of Haifa · Chaikin Institute for Geostrategy
Ami Pedahzur
Professor of Geostrategic Studies
Researching terrorism, political extremism, special operations forces, and the changing character of warfare.
"Understanding political violence requires moving beyond headlines to the networks, ideologies, and strategic calculations that drive it."
About
My research asks a central question: how do states and non-state actors use organized violence, and what does that tell us about power, ideology, and strategy? Over three decades I have studied suicide terrorism, religious and ethno-nationalist extremism, and the evolution of special operations forces — work that has taken me from IDF archives to field interviews across three continents.
I direct the Chaikin Institute for Geostrategy at the University of Haifa and hold a professorship in Geostrategic Studies. Before returning to Israel I spent two decades at the University of Texas at Austin, where I built an interdisciplinary research program at the intersection of security studies and computational methods. My books have been published by Oxford University Press, Columbia University Press, Polity Press, and Routledge.
Research Fields
- Terrorism and Political Violence
- Political Extremism
- Special Operations Forces
- Counterterrorism
- Social Network Analysis
- AI-Enhanced Research Methods
Current Work
Recent directions and active research themes
Special operations forces and political influence
Research on how elite military institutions shape doctrine, national security policy, and political leadership.
AI, data, and political violence research
Work on research design, social network analysis, and computational methods for studying security and conflict.
AI-Assisted Research Methods for the Social Sciences
Short notes on research repositories and public project sites that document how AI can support, but not replace, social-scientific interpretation.
This research program treats the AI workflow as part of the method. Each repository makes visible the model-assisted step, the prompt logic, the data structure, and the point at which human validation enters the pipeline. The goal is not automation for its own sake, but more transparent and reproducible research on political violence, security studies, and historical processes.
View GitHub ProfileManaging High-Volume Digital Sources in Political Research with Emerging Technologies
Jonathan Grossman, Orel B. Amano, and Ami Pedahzur circulated this January 2022 SSRN working paper before ChatGPT became publicly accessible. The draft argued that the central methodological challenge was no longer merely finding digital sources, but organizing high-volume, mostly unstructured material without losing context, source criticism, or human interpretation.
The paper became an early bridge to my later work on AI-assisted research methods: it framed personal knowledge management, annotation, linking, and retrieval as part of the research design rather than as private note-taking infrastructure.
Conceptualization: defining special operations forces
Research topic. This repository asks how scholars and doctrine writers have defined "special operations forces" across 69 definitions published between 1944 and 2018. AI-assisted extraction and coding help organize definitions around four dimensions: designation, organization, operators, and role/function.
Contribution. The project turns conceptual analysis into an auditable workflow. It shows that special operations forces behave as a family-resemblance category rather than a concept with one necessary core, and it uses multiple correspondence analysis to reveal variation across time and country.
Open repository →Historical Entity Tracker: multilingual entity resolution
Research topic. This repository develops a tool for tracing people, organizations, places, and events across multilingual historical corpora. It extracts entities and events from unstructured documents, grounds them in source quotations, and proposes cross-document links for review.
Contribution. The project is built around a conservative methodological principle: co-presence is a hypothesis, not a fact. LLMs generate confidence-scored candidate links, but a historian adjudicates them before they become evidence for process tracing or causal interpretation.
Open repository →October_7: reproducible data on unit deployment
Research topic. This repository hosts workflow, data, and analysis for studying the role of special operations forces during the October 7, 2023 attack. It includes location and unit datasets, descriptive infographics, inferential analysis, and a public project website.
Contribution. The project demonstrates how AI-assisted organization and analysis can support transparent research on contemporary historical events. The emphasis is on reproducibility: the data, analytical pipeline, visualizations, and human-reviewed outputs are connected in one open workflow.
Open repository →Settlements: a quantitative panel of territorial change
Research topic. This project assembles a settlement-level panel combining population histories, electoral results, socio-economic measures, built-up area, and geography across the West Bank, East Jerusalem, and the Golan Heights.
Contribution. A curated research workbook and reproducible pipeline connect source crosswalks, an analysis-ready database, and an interactive map. The data model, matching decisions, and analysis remain under active development.
Open repository →Books
Monographs and edited volumes
Articles and Chapters
Sourced from Google Scholar
Contact
Emailapedahzur@univ.haifa.ac.il
Google ScholarView Profile
ORCID0000-0002-0925-4993
University of HaifaFaculty Page
Curriculum VitaeOpen CV
GitHubResearch Repositories
OfficeEshkol Tower, Room 2306, University of Haifa