How the rankings work

Every chart is calculated by code from public data, not by editors or AI. The same data always gives the same ranking.

Two halves of the score

Korea buzz (KR) adds up 7 days of theqoo hot-board views and comments, Google Trends KR searches and Yonhap news mentions.

KR = 0.4·L(theqoo views) + 0.2·L(theqoo comments) + 0.2·L(Google Trends) + 0.2·L(Yonhap)

Global interest (GL) averages 7-day Wikipedia page views across 8 languages, so English Wikipedia can’t dominate the score.

GL = avg(L(Wikipedia views): ko, en, ja, zh, vi, id, fr, ru)

The final score gives the two halves equal weight.

Score = 0.5·KR + 0.5·GL

Every signal is log-scaled, then rescaled to 0–100 within that day’s candidates.

L(x) = log₁₀(1 + x)

Candidates are filtered by their Wikidata type: groups and singers for K-pop, actors for the actors chart, series for K-dramas. A comedian can’t land in the K-pop chart.
To be ranked, a candidate needs at least one Korean signal in the last 7 days and a Wikipedia article in at least one language.
Climbing 5 or more places within the top 30 counts as rising. NEW means the candidate wasn’t on the previous chart.
“Rising in your language” ranks the growth of Wikipedia views in that language (this week vs last week, at least 500 views).
Fan cheers are shown separately and never change the score. The weekly fan vote is a chart of its own.

theqoo hot board — post titles and view and comment counts only; posts and images are never copied.
Google Trends (Korea) — daily trending searches and their approximate volume.
Yonhap News — how often a name appears in headlines.
Wikipedia page views (Wikimedia API) — CC0 public domain.
Wikidata — names, types and Wikipedia titles, CC0 public domain.
Wikimedia Commons — photos, each under its own license and credited on every chart.

Our charts and data are free to reuse with credit under CC BY 4.0

YouTube view counts: the YouTube API Services policy (III.E.4.h) doesn’t allow ranking with its data.
Nate Pann and Bugs: their robots.txt asks bots not to collect.
Melon and Circle Chart: their robots.txt and terms don’t allow text and data mining.