Observed arrival · 2026-11-18
The Contact Trail Behind Online Drug Ads
A Korean-language interactive investigation traces illicit advertising across neglected websites and ordinary online spaces.
- For
- Readers following online crime, data journalism, and youth safety
- Worth noticing
- Four team members reviewed a sample of 1,832 posts; in a random 700-post check, human labels matched automated classifications 90.7% of the time.
Field notes
The team says it first located 118 poorly maintained sites, then excluded inactive sites and those without June posts, leaving 79 for a month-long collection. Posts were normalized for obfuscated spellings, grouped with a hand-built slang dictionary and a Naive Bayes model, then checked by four reviewers; the article reports 90.7% agreement with human labels on a random 700-post sample. A separate Google search across ten shopping platforms yielded 122 suspected ad results on seven platforms.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue