CoChat

The DOI matches its Crossref record; 79 of the 89 references it cites were found in a registry or an index; 66 of the 203 sources its claims cite were checked for support.

Tainted Love: a Systematic Literature Review of Online Romance Scam Research

Open access
Bilz, Alexander, Shepherd, Lynsay A, Johnson, Graham I · 2023 · Interacting with Computers

Abstract

Romance scams involve cybercriminals engineering a romantic relationship on online dating platforms for monetary gain. It is a cruel form of cybercrime whereby victims are left heartbroken, often facing financial ruin. We characterize the literary landscape on romance scams, advancing the understanding of researchers and practitioners by systematically reviewing and synthesizing contemporary qualitative and quantitative evidence. The systematic review establishes influencing factors of victimhood and explores countermeasures for mitigating romance scams. We searched 10 scholarly databases and websites using terms related to romance scams. The methodology followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines: a total of 279 papers were screened. One hundred seven papers were assessed for eligibility, and 53 were included in the final analysis. Three main contributions were identified: common profile features and techniques used by romance scammers, countermeasures for mitigating romance scams and factors predisposing an individual to become a scammer or a victim. Despite a growing corpus of literature, the total number of empirical or experimental examinations remained limited. The paper concludes with avenues for future research and victimhood intervention strategies for practitioners, law enforcement and the industry.

Overview

CoChat Summary

TL;DR

This PRISMA systematic review synthesizes 53 studies on online romance scams, covering scam profiling, victim/scammer factors, and countermeasures. It finds a growing but still limited empirical literature, with most work on scam progression and tactics and less on perpetrators, robust datasets, and usable interventions.

Key findings

  • The final review included 53 studies: 39 journal articles, 9 conference proceedings, 4 theses/dissertations, and 1 book chapter, published 2009–2023 with a rise in 2019–21; contributors came from 13 countries. Contributions were n = 29 on profiling romance scams, n = 14 on countermeasures, and n = 10 on factors of becoming a scammer or victim; qualitative studies (n = 27) were more prevalent than quantitative (n = 20).
  • Profiling work identified process models such as Rege’s three stages and Whitty’s five-stage model and Scammers Persuasive Model, plus linguistic stages (initial, pre-attraction, hooked), persuasion/repression tactics, isolation, and military-themed personas. Fraudulent profiles often misrepresented origin as USA (61% in Pan et al. 2010; 63% in Edwards et al. 2018), and scams originated from Nigeria (30%), Ghana (13%), Malaysia (11%), and South Africa (9%); reported losses ranged from £50 to £800,000, with medians of £1001–£10,000 per victim.
  • Influencing factors included active online dating engagement, desire for an international partner, and 1–5 years of internet experience; Saad et al. (2018) found well-educated, married Chinese and Malay women aged 25–45 had high victimization likelihood, while Whitty (2018) highlighted well-educated women aged 35–54. Scammer drivers included peer recruitment, poverty, unemployment, and low education/income (Barnor et al. 2020); gender risk findings were contradictory.
  • Countermeasures included machine-learning detection: Suarez-Tangil et al. (2019) reported 97% accuracy for an ensemble classifier; He et al. (2021) reported precision 90.5%, F1 0.857, and AUC 0.940 for DatingSec; Li et al. (2019) predicted gender from chat with 72% accuracy using keystroke dynamics and 64% using stylometry. Human-centred approaches included reverse image/internet searches, contextualized warning messages, peer support groups, and proactive policing, though datasets were scarce and technical tools were rarely end-user viable.

Methods

  • PRISMA systematic review searching 10 databases (Scopus, PsycInfo, PsycArticles, Medline, Web of Science, ProQuest, ACM Digital Library, IEEE Digital Library, ScienceDirect, PubMed) plus reference lists and a Google Scholar alert; search conducted October 2021 and updated April 2023.
  • Eligibility required empirical/experimental work profiling romance scams, psychological aspects, or countermeasures; accepted peer-reviewed journal articles, book chapters, conference proceedings, theses/dissertations, and grey literature, limited to English-language papers available through University subscriptions.
  • Two-stage screening in Covidence (title/abstract then full text) with a custom data extraction template; extracted metadata and study-specific findings, with all extracted data validated a second time.

Limitations

  • The paper states that advanced statistical processing was not feasible, direct comparisons were difficult due to limited homogeneity, many studies focused primarily on online crime rather than romance scams, and few influencing factors were statistically significant; gender findings were contradictory.
  • It also states that data are limited by under-reporting, ethical/legal challenges, scarce and repeatedly reused datasets, reliance on datingnmore.com and scamdigger.com that may not represent the general dating population, and technical countermeasures rarely viable for end-users; some studies did not fully disclose data sources.
  • Evident from the text: the abstract reports 279 papers screened while the results section reports 334 studies identified, and the search was restricted to English-language and University-subscription-accessible sources.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1093/iwc/iwad048

Cites

79 of 79 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

58 Valid · 7 Not valid · 107 Not checked · 78 Own results and interpretations

105 more claims · 24 with their cited sources checked

Create a free account to see all 250 claims and to verify the 4 references still open.

See all claims — free account
  • Probably not a substantial claim. 77
  • It may state two separate claims as one. 87

Create a free account to see these claims.

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

1 of 1

Citations

Cites

Show all 79 references

Research works printed but not linked: 20

  1. Action Fraud, 2023, Action Fraud
  2. Amazon, 2023, Amazon Rekognition Documentation
  3. Barclays, 2023, Romance Scams
  4. BBC, 2021, Romance fraud: ‘I wish I hadn’t given £300k to a man I met online, BBC News
  5. Buchholz, 2023, How the World Dates Online [Digital image]
  6. Covidence—Better systematic review management, 2022, Covidence
  7. Dating ‘n More, 2021, Dating ’n More
  8. Financial Ombudsman Service, 2023, Financial Ombudsman Service Warns of Increase in ‘Hybrid’ Scams
  9. Google Vision AI | Cloud Vision API | Google Cloud, 2023, Google. Google
  10. Graham, 2021, Automatic Detection of Fraudulent Dating Site Profiles Through Use of Reverse Image Searching
  11. Gregory, 2012, A sweetheart of a deal: how people get hooked and reeled in by financial scams, The Journal of Behavioural Finance and Economics, 2, 96
  12. HSBC, 2023, What's the Difference Between Fraud and a Scam?
  13. Momo, 2023, Momo
  14. Okcupid, 2023, Match Group
  15. Pizzato, 2012, User Modeling, Adaptation, and Personalization: 20th International Conference
  16. Scamwatch, 2023, Scamwatch, Australian Competition and Consumer Commission (ACCC).
  17. Smeitink, 2021, A Postmodern Love Story
  18. TinEye, 2023, Reverse Image Search. TinEye
  19. Yandex Search, 2023, . Yandex
  20. Yu, 2013, LibShortText: A Library for Short-Text Classification and Analysis

Cited by

See all 65

Metadata

Authors
Bilz, A., Shepherd, L. A., & Johnson, G. I.
Affiliations
Abertay University (GB)
Corresponding author
Shepherd, Lynsay A
Venue
Interacting with Computers
Published
2023-10-27
Volume
35
Issue
6
Pages
773-788
ISSN
  • 0953-5438 (ISSN-L)
  • 1873-7951
OpenAlex
W4388014937
Cited by
65
References
89
Citations by year
2026: 29 · 2025: 27 · 2024: 9
FWCI
2.69
Type
journal article
Language
en
Open access
Locations
  • Interacting with Computers (publishedVersion, cc-by)
  • Abertay Research Portal (Abertay University) (acceptedVersion, cc-by)
Funders
  • Abertay University
  • Scottish Funding Council
Topics
  • Social Sciences › Law › Cybercrime and Law Enforcement Studies
  • Physical Sciences › Computer Science › Information Systems › Spam and Phishing Detection
  • Social Sciences › Sociology and Political Science › Stalking, Cyberstalking, and Harassment
Keywords
romance scams, online dating, cybercrime, victimhood, countermeasures, victim intervention, financial fraud, online relationships

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-04, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-04 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Tainted Love: a Systematic Literature Review of Online Romance Scam Research. https://app.cochat.ai/papers/10.1093/iwc/iwad048
The paper
Bilz, A., Shepherd, L. A., & Johnson, G. I. (2023). Tainted Love: a Systematic Literature Review of Online Romance Scam Research. Interacting with Computers. https://doi.org/10.1093/iwc/iwad048

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.