LA NACION DATA (Argentina) - Open Data Journalism for Change

LA NACION DATA (Argentina) - Open Data Journalism for Change

Organisation: La Nacion (Argentina) (Argentina)

Publication Date: 04/14/2016

Size of team/newsroom:large

Description

LA NACION DATA website is not just a website, it´s a strategy, a project and a team. The strategy has to do with the commitment of using data to tell stories and expand the use of data, preferably open data, to activate demand of public information in a country still without FOI Law. // 1. THE year of ELECTIONS DATA. During 2015, Argentina had national and local elections including executive and legislative branches. We decided to push the limits in terms of: Innovate in data visualization giving full control to the user in cross platform interactive visualizarions and maps applications Build unique data set using Machine Learning and open collaboration, to have a unique level of granularity in the analysis of geolocated data. Then open and share this dataset. // 2. OPEN COLLABORATION TO MONITOR ELECTION TELEGRAMS IN PDF FROM POLLING STATIONS. We used crowdsourcing to monitor elections as the original telegrams from the polling stations were released after the primary elections and our national electoral system was suspected. We downloaded more than 70.000 PDFs and structured a from to build another database from scratch using Vozdata to input data from those telegrams and classifiy them to detect inconsistencies. After processing more than 14.000 using simultaneus validation by 2 we analyzed this data and reported on a front page story. The director of the National Electoral Office recognized this work in a press conference mentioning LA NACION and this evidence supported what others were claiming, a change in the electoral system. // 3. USING OPEN DATA TO MONITOR CITIZEN PARTICIPATION THROUGH CLAIMS. We used a data set of citizen claims in City of Buenos Aires with more than 2,5MM rows. As this is offered as open data so we just had to normalize and clean some data to report on four big topics Before the City elections we did maps and analysis about Garbage, Public Transportation, Social housing and security needs. // 4. READY OPEN DATA SERIES. Using data and opening it for daily reporting is showing context and evidence but also letting users REUSE this datasets in their own stories or analysis. Examples: dolar, euro or commodity prices, inflation (IPC) from various series, u$s reserves in Central Bank, real estate activity in Buenos Aires, monthly sales in automotive industry. // DATA AS EVIDENCE: 5. INDEC, The Machine of Lies. Multimedia scrollytelling format to tell the story about the scandal of false statistics that INDEC was lying about, specially regarding inflation (CPI) and poverty. 6. Argentina´s PRISONS System series // 7. USE OF DATA FOR SIMULATION: INTERACTIVE DATA CALCULATORS e. g. Income tax calculator (based in algorithm with new tax rises), Elections Ballotage simulator (based in General Elections data), Electricity prices calculator (based in new prices data just announced) LONG TERM PROJECTS - BUILD ONCE, USE, OPEN, UPDATE, REUSE 8. Congresoscope 9. Open Declaration of assets, 10. Open Data Catalog.

What makes this project innovative? What was its impact?

Open collaboration through civic marathons with more than 6 Universities, NGOs and individuals // Structuring citizen participation through an open collaboration platform (Vozdata / Crowdata) to observe PDFs and classify them, then reporting and opening this data as evidence (giving back to the community) // Use of Machine Learning to optimize the deduplication of addresses that were suggested by proximity of three sources joined in one application // Using technology to centralize and serve multiplatform graphics sourced by the same data flow // Saving the data for future investigations, keeping it safe, transforming it to build layers of knowledge for our journalists // Keeping datasets updated from PDFs, some of them compiling 10 years of exclusive data, and opening these datasets as well // Opening daily data while reporting, the data daily series.

Technologies used for this project:

D3js, underscore, RequireJS, SVG, html5, css3, gulp, bower, Python , Django, PostGress, // Jquery, Hammer.js, Ruby, RoR, Angular.js, Handlebars.js, Xapian // MS Sql Server, Visual Basic for Applications // Excel, Google Spreadsheets, Open Refine, Junar Open Platform, Carto,db SQL API, Leaflet Draw Plugin, Mapbox Studio // PDF To Excel, Abby Pdf converter, Tabula.technology, Iport.io, NitroPDF, Omnipage, // DocumentCloud, Timeline.js // Machine learning process using csvlink
Follow this project
Wait
Comment

Comments (0)

You have to be connected to contribute

You have to be connected to follow

Leave this project and no longer be informed about this project

By joining this project, you will be informed by email when an update or a new contribution is posted on the website.

Thank you for your active participation !
Best,

The GEN Community Team