News sentiment API: scoring urgency and political lean

Classic sentiment — positive, negative, neutral — tells you little about a news event. Is a "negative" article a market-moving crisis or a routine grumble? NewsAgent Data replaces polarity with two signals built for news: urgency and political lean.

Why polarity falls short for news

Sentiment models were built for product reviews and social posts. On news they collapse very different events into the same bucket: a fatal disaster and a critical op-ed can both score "negative." What you actually want to know is how important is this and from what stance is it told.

Urgency scoring (0–10)

Every article is scored for urgency by a deterministic 300+ pattern engine — mass-casualty language, sanctions, central-bank moves, military escalation, disasters. A threshold means the same thing every day, so min_score=7 reliably surfaces what matters and filters the noise.

Political lean (9 categories)

Each item is classified across state, official, centrist, liberal, conservative, nationalist, opposition, tabloid and neutral — so you can measure framing and compare how different camps cover the same event, programmatically.

curl -H "X-API-Key: YOUR_KEY" \
  "https://api.newsagentdata.com/v1/feed?min_score=6&political_lean=opposition&days=3"

Comparing narratives

Because items covering one event share a cluster_id, you can pull a single story across state vs. independent vs. Western sources and contrast the lean distribution — a far richer signal than one polarity number. More in our guide on measuring political bias.

How the scores are built

The scoring is rules-based and documented, not a black box — read the methodology for exactly how urgency and lean are assigned, and the API docs for every field.