Last Saturday saw the 2nd installment of GovCamp, this time titled ‘Hack the Government’, in Amsterdam. A full day long both civil servants and coders discussed different issues around opening up government data. At the same time the coders that were present created applications right there and then, based on open government data. The day was put together by my friend James Burke and Lex Slaghuis, and with loads of much appreciated help by Vincent Lindeboom and Edial Dekker. I started the day with a short introduction sketching the general landscape around open gov data, based on our experiences in the project described in the previous posting, and facilitated the audience in building the program BarCamp-style on an impromptu-wiki-wall.
Citizens and coders need to take more of a constructive activist stance
Like I remarked last year, it is one thing to complain about government doing things ‘wrong’ and ‘not getting it’, and treating government as one organizational body, but quite another to actually formulate what you want and what you are prepared to do yourself to get there. This is regardless of the fact that there are civil servants who don’t ‘get it’ at all, and are making mistakes, because when you don’t reach out to help them ‘get it’ they will never know they’re not living up to your expectations in the first place. This is regardless of the fact that we often experience stonewalling when asking for more government transparency, because if we don’t do whatever already is possible within our own sphere of influence we will never get to the point where the still dominant culture of default opaqueness is changed. Just saying ‘yes, but it’s the law’ is not enough and will not work change, you have to be prepared to act based on it. As large scale political pressure to make this a priority is lacking (unlike e.g. the US currently, or the UK), we need to build pressure ourselves.
Civil Servants need to reach out
At the same time it is key that civil servants who are serious about open gov data let the ‘outside’ know what they need to move things forward, what we can do to help. Luckily, even in the midst of a lot of confusion over language and what is and what is not a governmental task, last Saturday was a platform for civil servants and coders to find a way to collectively move forward.
I was very pleased to see that a significant group of government employees decided to participate. Government ministries (e.g. Interior, Education, General Affairs) were represented, as well municipalities (e.g. Vught, Utrecht, Den Haag), and different parts of governmental bodies dealing with transparency, ICT or other topics (e.g. ICTU).
Andre Herbrink of Min of Education on open data (left) and Twitter wall back channel (right)
One of the presentations given, on ikregeer.nl
Alper’s presentation on scraping.
Not just talk, also walk
We not only talked about open data and discussed existing efforts and examples, we walked the talk as well. After Alper giving a short introduction on how to scrape data when the source is not directly available, or there is no API, coders got to work on different ideas and apps that reuse open (or in some cases not so open yet) government data.
Some of the things that people worked on:
An API for the RDW website and database available from OnzeData.nl (‘our data’), allowing access to information about cars based on their license plates. Created by Christoph Kempen en Manfred Zielinski. Update: RDW, the government agency involved, is currently actively blocking the API. Proving how slow going this process of transparancy is.
An Android application taking the information about cars available at the RDW, based on the user entering a license plate number. This info is available through a website, but with this app also more directly and quickly on your mobile phone. Built by Ronald v.d. Lingen.
Polirazzi, a website that takes a politician’s name and then gives you an overview of any newspaper articles mentioning that politician. Built by Breyten Ernsting, based on an API available at IkRegeer.nl (‘I govern’) to get the list of current members of parliament and party affiliation.
Discussing meaningful data visualization (l), Breyten and Ronald coding (r)
A way for investigative journalists to detect changes in PDF files published by government. Sometimes small but crucial pieces of info get augmented, changed or edited in already public documents without it being noticed and under the same url and document title. A small group worked on a way of versioning these documents.
An idea to add RIVM environmental data to an iPhone app, and combine it with existing rain radar apps. Reminiscent of the ‘smog alarm’ website we built as an example earlier.
A open version of the database of the Chamber of Commerce ‘who earn more money exploiting their database with your data than the average successful Russian spammer outfit’, called OpenKvK. With a demo of life scraping and searching at the end of the afternoon. It also allows automatic creation of a ZIP code database. This is an extremely important dataset (for all kinds of location aware services), that is currently owned by TNT to everybody’s chagrin and not in public hands (as an undesirable result of privatizing the mail). Watch OpenKvK.nl for when it’s ready (search form not active yet).
Do you know exactly when garbage gets collected in your street? And old paper? Garden refuse? Pascal van Hecke, Hanno Lans and Menno Sman worked on an idea to create an easy site giving you all the info you need based on your zip code and house number. Most intriguing part of their idea was to me their suggestion to crowdsource the needed data scraping: have a few people in each municipality scrape the data relevant and useful to them, and in that way build up national coverage and completeness.
During a session (l), lunch time conversations (r)
I may have missed some other efforts but, am happy to add more if you point out my omissions. I do have one regret though, and that is noone worked with the data of the Ministry of Education. They were present and could have helped explain jargon language etc being used in their sometimes hard to understand API, also they were really eager to see people work with their data, and actively asking for feedback and input.
(More pics in my Flickr stream)
A short video impression (in Dutch) by Elmine Wijnia: