HAVEN
Hospitality research / Published · 2026
HAVEN
Helping hotels detect human trafficking together without sharing guest data.
I coauthored HAVEN: Helping Hotels Detect Human Trafficking Together Without Sharing Guest Data with Saeed Vayghan and Mehmet Erdem. The paper is published in ICHRIE Research Reports, Volume 11, Issue 6, Article 2 (2026).
HAVEN stands for Hospitality Anti-trafficking Vigilance and Early-warning Network. The research explores how hotels could learn collectively from booking patterns while keeping individual guest records at each property.
Research translated into a working concept
The paper presents a conceptual framework using federated learning and differential privacy, accompanied by a web demonstration across 12 simulated hotels. It examines hotel operations, human review of alerts, implementation readiness and the ethical questions that a future pilot would need to address.
Published in 2026. The demonstration uses synthetic data. Real-world detection accuracy, fairness and false-positive rates require further evaluation with hotel partners and appropriate ethical oversight.
Research record
Question, approach and contribution
Research question
Could hotels learn collectively from booking patterns while retaining guest records at each property?
Approach
A conceptual framework combining federated learning and differential privacy, illustrated through a web demonstration across 12 simulated hotels. All demonstration data is synthetic.
Contribution and collaboration
Coauthored by Mahboubeh (Mia) Cheraghian, Saeed Vayghan and Mehmet Erdem. The research translates technical concepts into hotel operations, implementation readiness, human review and ethical considerations. Technical implementation is credited to the research team; individual ownership of every component is not claimed.
What the demonstration can establish
It illustrates the proposed architecture and workflow. It does not establish production detection accuracy, fairness or false-positive performance. A real-world pilot requires hotel partners and ethical oversight.
Contribution and implications
The synthetic-data demonstration illustrates a proposed coordination architecture. The contribution is a framework for evaluating collective learning, property-level data control and human review in hospitality.
It does not show real-world detection accuracy. Future pilots must evaluate false positives, fairness and operational feasibility under ethical oversight.
Published article
Read the paper
Cheraghian, M., Vayghan, S., & Erdem, M. (2026). HAVEN: Helping Hotels Detect Human Trafficking Together Without Sharing Guest Data. ICHRIE Research Reports, 11(6), Article 2.
DOI: 10.61701/680143.777