Transparency
Data Sources & Methodology
Exactly where TodayInPulse's information comes from, and what is calculated vs. AI-generated. We only use official public APIs — we never scrape websites.
Last updated: June 2026
News discovery
Headlines and article links are discovered via the GDELT Doc 2.0 API (a free, public global news index). GDELT provides the URL, title, publisher domain and first-seen time — it does not provide article bodies, so TodayInPulse never reproduces full text.
Recognized publisher coverage
For recognized major outlets (e.g. BBC, CNN, Fox News, Reuters, AP) we run domain-targeted GDELT discovery to improve consistent coverage monitoring, and we surface a Recognized Publisher badge when such an outlet is among the sources. We do not scrape these sites, and direct/licensed feeds stay disabled where commercial-use restrictions apply. Recognition identifies the outlet and affects coverage monitoring only — it never affects ranking and is not a check of the story's facts.
Social sources
Social posts come from official public APIs: YouTube Data API (most-popular videos) and Bluesky (public top posts). We store only real, publisher-provided engagement metrics (views, likes, comments/replies) and link every post to its original platform.
Open Web sources
Non-social attention signals come from Hacker News (official Algolia API: points/comments) and Wikimedia (official pageviews API — a daily aggregate, not real-time velocity). These are attention signals, kept separate from the social ranking.
Calculated TodayInPulse metrics
Pulse Score, TodayInPulse Social Score and trend momentum are TodayInPulse-calculated (0–100) measures of attention/velocity, normalized within each source. They are not official metrics from any platform and are not a measure of truth or quality.
AI-generated content
Neutral story summaries and category labels are generated by an AI model (GPT-5.4) and are constrained to the real headlines and source list shown on each story. The model is instructed not to add facts, quotes, numbers or publishers beyond those sources, and a summary can never appear without an underlying source record. As with any language model, wording can occasionally be imperfect or miss nuance — so every summary links to its original sources for you to verify, and summaries describe coverage, not verification of the facts themselves.
Corroboration vs. publisher recognition
These are two separate things, and neither is a fact-check of a story's claims:
- Unverified — only one independent report, or not yet enough independent corroboration.
- Multiple Sources — more than one genuinely independent report (wire and affiliate copies of the same report are collapsed, so they don't inflate the count).
- Multiple Independent Reports — three or more genuinely independent reports corroborate the coverage. We use this phrase rather than "Confirmed" on purpose: it describes how widely the story is independently reported, not that we have verified the underlying facts.
- Recognized Publisher — an outlet whose domain matches TodayInPulse's recognized-publisher registry is among the sources. This identifies the outlet; it says nothing about whether the story's claims are true.
Every count you see reconciles with the source rows on the story — expand Sources to inspect each counted outlet and open the original article.
Live vs. Demo data
Live content is clearly badged Live; any placeholder/sample content is clearly badged Demo Data. Demo content is realistic but fictional and never mixes into live rankings.
Corrections
Spotted a wrong grouping, category, trend link or broken source? Use the Report an issue button anywhere in the app, or email hello@todayinpulse.com.
Questions? hello@todayinpulse.com · Data & Methodology