Editorial Process
How Apiar Data selects, matches, and publishes news briefings: our editorial standards, data sourcing, and transparency commitments.
What we publish
Apiar Data News Briefings are concise topic summaries that pair a current news cluster with a relevant statistic from our dataset catalog. Each briefing answers a single question: what does the data say about this event?
We do not publish breaking news, opinions, or forecasts. We publish the institutional data that frames the context around events that are already being widely covered.
How topics are selected
Topics are identified using a TF-IDF clustering pipeline applied to a curated set of RSS feeds spanning over 2,400 international publications. Clusters must cross two thresholds before becoming a briefing:
- Coverage threshold — the topic must appear in at least three independent sources in a 24-hour window.
- Data match threshold — our system must identify at least one statistical dataset in the Apiar Data catalog with a meaningful connection to the topic’s primary entities (countries, organisations, economic variables).
Topics that meet both thresholds are matched to a lead statistic and up to five supporting datasets. The match uses entity normalisation and domain tagging against our catalog metadata.
Data sourcing and accuracy
All charts in news briefings are sourced directly from the linked statistics pages, which themselves reference their original sources (World Bank, national statistical agencies, OECD, BIS, etc.). We do not produce original statistics — we contextualise published data.
Specific figures cited in briefing text (e.g. percentage changes, dollar amounts from news articles) are sourced from the linked articles listed in each briefing’s source section. We do not independently verify these figures beyond their appearance in multiple credible publications.
Attribution and licensing
News clustering is based on publicly available RSS feed data. Source publications are credited in each briefing’s source list and linked wherever URLs are available. We do not reproduce article text; only headlines and metadata are used in the clustering process.
Statistical data is sourced under the terms described in our Data License.
What we do not do
- We do not take editorial positions on geopolitical events.
- We do not generate news or create original reporting.
- We do not profit from the news articles we link to.
- We do not share or sell reader data. See our Privacy Policy.
Contact
If you believe a briefing contains an error in the statistical data, or if you are a publisher and have concerns about attribution, contact us at daniel@davia.ventures.