Longitudinal data gathering andanalysis of Dark web marketplaces & Analysis of cannabis retail on the Dark web and market impact of legalization
Bibliographic record
Abstract
I:\nDark Web marketplaces have been in operation for more than a decade, and they are host to a vast number of retailers and customers who exchange illegal goods and services. Leveraging the anonymity of the Tor network and the resilience of cryptocurrencies against censorship and audit, these marketplaces have remained an enduring nuisance for law enforcement and prosecutors. Trends and metrics on these marketplaces are a novel source of information, but it is a non-trivial undertaking to access, retrieve and systematize this data.\n\nFirstly, this paper documents our design, implementation and operation of a scraping software which accomplishes this task. The software consistently scraped marketplaces within 24 hours and reliably subverted marketplace measures designed to evict bots. We scraped three marketplaces, Empire Market, Cryptonia Market and Apollon Market, and parsed data from ca. 180 000 unique listings over a period of 150 days. Additionally, we parsed another 260 000 listings from offline crawls of Dream Market in the period from January 2014 to November 2019. Secondly, based on our collected data, we present quantitative analyses which characterize economic aspects of the Dark Web marketplaces. We examine product types, vendors, prices, quantities and more, and cross-aggregate these entities by time, geography and other attributes, revealing many trends and metrics for both individual marketplaces and the industry of Dark web retail at large.\n\nII:\nDark web marketplaces have been in operation for more than a decade, and they are host to a vast number of retailers and customers who exchange illegal goods and services. Cannabis is one of the most sold items on the Dark web marketplaces, and one of lawmakers' main goals of legalizing cannabis is to marginalize this illicit industry. \n\nUsing recently obtained data from the marketplaces, we explore characteristic properties of the cannabis market and analyze the effects of legalization. Using natural language processing techniques and leveraging geographical attributes in our data, we have been able to calculate unique average per-gram prices of cannabis by country, enabling a comparative, quantitative evaluation of individual cannabis markets.\n\nWe have studied the impact of the Canadian Cannabis Act and the Australian Drugs of Dependence (Personal Cannabis Use) Amendment Bill 2018 on the Dark web cannabis market. During the first 18 months after the Canadian law was enacted in October 2018, Canadian prices dropped by 57 % and relative sales volume of cannabis increased by 26 %. We did not observe any significant impact of the Australian law, probably because this law was relevant only for Australian Capitol Territory, and our data does not allow us to study this area separately from the rest of Australia.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.011 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.005 | 0.005 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".