About the data

These are the official sources that feed the platform:

• Congress of Deputies open data - congreso.es/es/opendata. Deputies, initiatives and plenary votes. CC-BY 4.0 licence.
• Congress initiative search - busqueda-de-iniciativas. Series 162 (non-legislative motions) and 173 (motions). Same licence.
• Official texts - the Official Journal of the Cortes Generales (BOCG), where we read the text of each bill and motion, and the BOE, to know when an approved law comes into force.
• Chamber seating - the official seat map published at congreso.es/es/hemiciclo. The image and coordinates belong to Congress; we only show where each deputy sits.
• Deputy profiles - public pages on the Congress website. From them we take the official photo, parliamentary code and year of birth. Never private data (personal phone, address, family).
• Wikidata and Wikipedia - the link to each deputy's article and a short extract, under its licence (CC BY-SA).

Documented processing pipeline:

1. Regular download of the official datasets. A scheduled job fetches new votes every 4 hours, and deputies and initiatives once a day.
2. Normalisation into our database, with a documented schema. No computed figure is made up: everything comes from an identifiable column in the source.
3. AI topic classification. Each initiative is classified by topic (17 editorial topics plus the 17 SDGs of the 2030 Agenda) with a European language model (Mistral). The taxonomy is closed: the model picks from the defined topics and never invents one. Accuracy varies; unclassified initiatives say so.
4. AI plain-language summaries. Since the official text is legal language, we write a plain-language summary in Catalan and Spanish. The model has strict instructions not to give opinions: if a summary comes out with evaluative language, it is not published.
5. Aggregate metrics (cohesion, attendance, agreement between groups, approval rate by topic). All are computed with formulas documented in the source code, with no hidden weightings.

• Plenary votes: checked every 4 hours. Congress publishes results 24 to 48 hours after the sitting, so they usually arrive within 1 or 2 days.
• Initiatives: every morning. Most are published on the day they are registered or the day after.
• Sitting deputies: every morning. Picks up replacements and group changes.
• Seat map: by hand, when we know seats have been reassigned.
• AI classification and summaries: automatic, several times a day for new initiatives. The project pays the cost.

Hola Política is civic infrastructure, not an opinion platform. These are the lines we draw:

• No “best” or “worst” deputy rankings.
• No automatic rating of how good a law is.
• If we publish a group's high cohesion, we publish the low too, side by side.
• If we publish “who approves most” on a topic, we also publish “who rejects most”, in the same widget.
• No comments, reactions or parallel polls: the tool is a mirror, not a megaphone.

If at any point you see content that looks editorial or one-sided, it's a bug. Let us know.

• Linking votes to initiatives: Congress gives procedural steps (162 non-legislative motion, 173 motion...) different file codes from the underlying laws (121, 122...). When the code matches we link them; when it does not, the vote has no topic. We are working on matching them by text.
• Senate: not covered yet. The Senate does not publish individual votes in an open, structured format. When it does, we will add it.
• History: votes from the 10th legislature (December 2011) to the current one. A few old sittings are missing because the official source does not serve them.
• Language: the original content is in Spanish. Titles and summaries are translated with AI; the official text is always one click away.

If you find a data error (mislabeled vote, misclassified deputy, outdated profile) or a privacy issue, write to us. We resolve each case with traceability: we identify the source, fix it, and if the error came from the official source we forward the feedback.

Contact email: dades@holapolitica.org

• Source code: EUPL-1.2 (European Union Public Licence). Available at github.com/Pardo24/holapolitica.
• Data: CC-BY 4.0. You may reuse it crediting Hola Política and the original source (Congress of Deputies).
• Official photos: owned by the Congress of Deputies. We show them on the same terms it publishes them on its portal.

We process data on people who exercise public functions under the public-interest legal basis for transparency and journalistic purposes (Art. 85 GDPR, Art. 28 LO 3/2018, Spanish Transparency Act 19/2013). We never show private data, only mandate-related information.

If you believe the treatment of public data about your parliamentary activity is inappropriate in our context, write to us and we evaluate each case individually.