Skip to main content

Forensic footprints in digital love: Unveiling romance scams with Maltego and machine learning

Vedhanayagam, Priya ; Singh, Moulik ; Dadhania, Arora Preksha ; Ikram, Sumaiya Thaseen (2026) — AIP conference proceedings

DOI: 10.1063/5.0298575
Type:   Conference Paper
Country:   Global
Tags:   romance scams, online dating, OSINT, digital forensics, Maltego, machine learning, phishing detection, blacklisting, romance scam detection, logistic regression, email validation, phishing websites, alerts, website blacklisting, IPQS

Synopsis (AI-Generated)

The study compiles data from open-source scam-reporting sites and Maltego OSINT extraction. Maltego transforms gather website, profile, URL, email, DNS, SSL, hash, IPQS, and linked-account information. Emails and links are cleaned and represented with TF-IDF features; logistic regression classifies suspicious entities and assigns suspiciousness scores. The paper states that romance scams can cause financial, reputational, and emotional harms with long-term psychological effects. The reported experimental evaluation achieved 85% accuracy, 88% precision, 80% recall, and an F1 score of 0.84. The authors state that the proposed algorithms are not infallible and that their effectiveness and accuracy depend heavily on the quality and relevance of input data.

Identified Gaps (AI-Generated)

The paper identifies a gap in using forensic tools for in-depth romance-scam investigation. It also notes that existing detection methods can have high false-positive rates and may not adapt well to evolving scam tactics. The related-work review describes romance-scam research as limited, fragmented, and inconsistent, with constrained access to victim–scammer data and a predominance of qualitative studies.

Methods (AI-Generated)

The study compiles data from open-source scam-reporting sites and Maltego OSINT extraction. Maltego transforms gather website, profile, URL, email, DNS, SSL, hash, IPQS, and linked-account information. Emails and links are cleaned and represented with TF-IDF features; logistic regression classifies suspicious entities and assigns suspiciousness scores. The system alerts users and adds high-risk sites to a blacklist. Reported evaluation metrics are 85% accuracy, 88% precision, 80% recall, and F1=0.84.

Limitations (AI-Generated)

The authors state that the proposed algorithms are not infallible and that their effectiveness and accuracy depend heavily on the quality and relevance of input data. The approach relies on data extracted from external websites, Maltego transforms, and validation services, so incomplete or inaccurate extraction can affect results. Although performance metrics are reported, the supplied text does not describe dataset size, class balance, ground-truth construction, or external validation.

Future Work (AI-Generated)

The paper proposes prospective enhancements to further fortify the model and extend preventive security measures. Grounded in the discussion, future work should improve data extraction quality and relevance, strengthen adaptation to evolving scam tactics, and further develop the alerting and blacklist functions for online dating contexts.

Related Research Insights

See how this publication connects to RSRC's living evidence syntheses through current citations and research-topic mapping.

AI-Generated Content Notice

The synopsis and research notes on this page were generated with AI from available publication information and, when available, the uploaded paper text. They may contain errors, omissions, or interpretation issues. Readers should follow the DOI or source link, review the original publication, and make their own judgment about the content.

Found a possible error? Request a correction.