Dealings on the Dark Web: An Examination of the Trust, Consumer Satisfaction, and the Efficacy of Interventions Against a Dark Web Cryptomarket

Abstract Objective. The overarching goal of this thesis is to better understand not only the network dynamics which undergird the function and operation of cryptomarkets but the nature of consumer satisfaction and trust on these platforms. More specifically, I endeavour to push the cryptomarket literature beyond its current theoretical and methodological limits by documenting the network structure of a cryptomarket, the factors which predicts for vendor trust, the efficacy of targeted strategies on the transactional network of a cryptomarket, and the dynamics which facilitate consumer satisfaction despite information asymmetry. Moreover, we also aim to test the generalizability of findings made in prior cryptomarket studies (Duxbury and Haynie, 2017; 2020; Norbutas, 2018). Methods. I realize the aims of this research by using a buyer-seller dataset from the Abraxas cryptomarket (Branwen et al., 2015). Given the differences between the topics and the research questions featured, this thesis employs a variety of methodological techniques. Chapter two uses a combination of descriptive network analysis, community detection analysis, statistical modelling, and trajectory modelling. Chapter three utilizes three text analytic strategies: descriptive text analysis, sentiment analysis, and textual feature extraction. Finally, chapter four employs sequential node deletion pursuant to six law enforcement strategies: lead k (degree centrality), eccentricity, unique items bought/sold, cumulative reputation score, total purchase price, and random targeting. Results. Social network analysis of the Abraxas cryptomarket revealed a large and diffuse network where the majority of buyers purchased from a small cohort of vendors. This theme of preferential selection of vendors on the part of buyers is repeated in other findings within this study. More generally, the Abraxas transactional network can then be viewed as set of transactional islands as opposed to a large, densely connected conglomeration of vendors and buyers. With regard buyer feedback, buyers are generally pleased with their transactions on Abraxas as long as the product arrives on time and is as advertised. In general, vendors have a relatively low bar to achieve when it comes to satisfying their customers. Based on the results of the sequential node deletion, random targeting was found to be ineffective across the five outcome measures, producing minimal and a slow disruptive effect. Finally, these strategies are based on a power law where a small percentage of deleted nodes is responsible for an outsized proportion of the disruptive impact. Conclusion. As with all applied research examining emergent phenomena, this thesis lends itself to a more refined understanding of dark web cryptomarkets. While the results and conclusions drawn from these results are not perfectly generalizable to all cryptomarkets, they should serve to inform law enforcement on the dynamics which undergird these markets. To this extent, a sombre consideration of trust, consumer satisfaction, and tactical effectiveness of interventions is a necessary step towards the development of more effective countermeasures against these illicit online marketplaces. For law enforcement to be more effective against cryptomarkets, it is advised that an evidence-based approach be taken.

Name dc:type.qualificationname

Doctor of Philosophy (PhD)

Level dc:type.qualificationlevel

Doctoral

Grantor dc:publisher.institution

University of Cambridge

Year dc:date.issued

2021

Author dc:creator

• Harinam, Vincent

Advisor dc:contributor.advisor

• Campana, Paolo

• Dark Web
• Criminology
• Social Network Analysis
• Criminal Network
• Adaptive Computer Simulation
• Text Mining

Licence

• https://www.rioxx.net/licenses/all-rights-reserved/

Language dc:language

eng

DOI dc:identifier.doi

https://doi.org/10.17863/CAM.78755

OAI identifier oai:identifier

oai:www.repository.cam.ac.uk:1810/331308