povPR

Methodology

The idea is simple: see the full coverage, not just one headline.

povPR compares how different Puerto Rico sources cover the same story. Instead of slapping a PNP, PPD, or PIP label on each article, we look at the coverage, the sources, the possible gaps, and the editorial framing.

Story clustering

We gather articles that appear to cover the same topic using headlines, themes, mentioned names, and how close they are in time. The MVP prefers to keep stories separate rather than merge distinct events.

Source profiles

Each source has context about ownership, coverage model, editorial role, and profile confidence. That context is not used as an automatic label for every article.

Coverage gaps

We flag when we do not find coverage in the active sources we track. That can be due to timing, editorial focus, ingestion limits, or the fact that the source does not cover that kind of story.

Micro-audits

On sensitive topics, we compare wording, quoted voices, and framing without forcing partisan labels.

Ground rules

  • The feed is organized by stories, not by individual articles.
  • Source profiles have to be documented and open to public review.
  • A coverage gap means we have not found a record yet; it is not proof that an outlet ignored something on purpose.
  • Micro-audits compare framing, vocabulary, and whom outlets quote on sensitive topics.
  • Politics in Puerto Rico is too tangled to reduce every story to three labels.
  • Every detected cluster appears in the public feed, including early and low-confidence ones, instead of being hidden.
  • The full extracted text stays private; the public app uses metadata, excerpts, and links to the original outlets.

What a blindspot means

A blindspot in povPR means “we have not found coverage in these tracked sources yet.” It does not mean censorship, a cover-up, or editorial intent. Making a claim like that would require external evidence and human review.

We group tracked outlets into broad-reach commercial and broadcast sources on one side and independent, investigative, or ideological sources on the other, then look at how a story is split across them. The strongest label appears when a civic story is carried by the independent press but by none of the major outlets most people read. A second case flags widely-reported accountability stories that the commercial mainstream covered but no independent or watchdog outlet has.

Every gap is only called after a short window that lets late coverage arrive, and stories are re-checked as they mature. Even then, the explanation stays cautious.

Micro-audits

Micro-audits are reserved for sensitive clusters: status, corruption, public contracts, the Fiscal Control Board, LUMA, elections, privatization, and other civic topics. They compare wording, quoted voices, and framing, but they do not force PNP, PPD, or PIP labels when the article’s evidence does not support it.

Editorial disclaimer

Summaries, coverage notes, and micro-audits are generated automatically. Use them as signals to orient yourself, not as final truth. The full editorial and AI disclaimer, along with our terms of use, lives in our Terms of Service.