Well hello there, welcome to Dieren Theater!

Right now you are probably asking yourself a legitimate question, what is Dieren Theater? Dieren Theater is a project to offer to anyone structured and open data extracted (in fact scraped) from the website of the Belgian parliament (also knows as La Chambre or De Kamer).

Again, you might are asking yourself "Why? What is the meaning of that?". Well in fact, open data is one of the first step for us citizens to appropriate ourself political informations (political as in « the life of the city »), the way it is treated and displayed, to move from passive consumer to active creator and to get involve in this democratic process.

Sound too theoretical? Lets me show you some example:

• Nos Députés.fr and Nos Sénateurs.fr (both in french), two website from Regards Citoyens (a citizen group) that allow us to see the activities of the French deputies and senators, the subject they are the most working on, their interventions and many others informations. Those informations wasn't displayed in that what at all on the French parliament website.
• The Political Memory (English) which is a tool that allow an organization (here La Quadrature du Net) to put online the voting recommendations they have been sending to elected representatives (here deputies of the European Parliament and French deputies) and show who is following their recommendations the best. This allow here for example to see which elected representatives is the more for Fundamental Freedoms, Net Neutrality and other in the sens that defend la Quadrature du Net.
• Pippi Longstrings (English) which is a tool to do text analysis on European Laws related documents and to see for example (but not only) which documents have text in common
• Parltrack (English) which allows to track the legislative work of the European Parliament especially by following dossiers and/or committees.
• An interesting visualisation (French) by Baptise Coulmont which display the relations betweens French deputies based on how often they have been working together (by proposing laws).

None of those would have been possible without data and none of this has yet happen (in my knowledge) in Belgium. So Dieren Theater is an attempt to build the first step by providing data to everyone and therefor allowing everyone to build similar tools or analysis.

It is interesting to notes that (except Baptise Coulmont works) all of these website are Free software (as in speech).

For the moment this project can be considered as an alpha, not because it is unstable be because it is very young, that a part of the data is missing (not already scraped) and that the available data (you have a list bellow) will probably needs polishing.

Roadmap (aka thing that I think I'm going to do)

• get more data, obviously
• but sooner having an intelligent automatic update strategy of data to avoid doing a sequential parse every night
• storing an history of the modifications of every data
• ...
• PROFIT

I'm looking for feedback on the data from people that are trying to build or play with it. Don't hesitate to contact me or to open a ticket on the bug tracker.

In a long term it is totally possible to think of parsing other similar websites than (lachambre|dekamer).be (like the senate website).

Info those links point you to a web interface that gives you an idea of the available data and their format, to get the actual data use the API bellow.

Warning small differences between the data displayed on the web interface and the data get from the API is possible. If you notice this situation, please report a bug.

As requested by some persons, here are dump of each collections of the database. All data are in json since it's coming from mongoDB. For bandwidth reason I'm compressing them using xz:

• Deputies
• Commissions
• Commissions membership
• Law projects/texts
• Written questions
• Annual reports

Okay, lets start funny things, how to get data!

Everything starts here:

http://www.dierentheater.be/api/v1/?format=json

And in a more convenient way (in a shell if you are using an unix based operating system):

curl "http://www.dierentheater.be/api/v1/?format=json" | python -m json.tool

The API is self describing so I won't write a lot here about the data in it. If you want to have and idea of what is available just browse it via the links in the previous section.

An interesting possibility of the API is that you can add the GET argument "lachambre_id" to the url to query an exact data like this:

curl "http://www.dierentheater.be/api/v1/document/?format=json&lachambre_id=2005" | python -m json.tool

Always use this way a querying an individual document whenever possible if you want to build something that is supposed to last, the id field (coming from mongoDB) isn't reliable at all for the moment.

You can find tastypie API documentation here

Available formats are json, xml and yaml (just replace json with xml or yaml in the url).

I hope you'll have fun playing with it!

This website is build in Python, using django-nonrel, a fork of django to use mongoDB instead of a SQL database. Parsing is done using the very awesome BeautifulSoup with long sessions under IPython. The API is done with a patched version of tastypie-nonrel. And I use boostrap to make you believe that I know how to do a pretty website.

So you want to help? Well thanks a lot for your proposition!

So, how can you do this?

• well, the first obvious way to help this project is to spread the love, this project doesn't make any sens if nobody builds things out of it so talk about it to everyone you know that is likely to do so!
• the second obvious way to help is to build cool things out of it! Send me an email if you do so, this will boost my motivation.
• the third way is to get involve technically in the project, the used technologies are listed above and the code is here. Fork the depot and send me a pull request. If you are planning on doing something a bit big it's a very idea that we talk about it before to avoid you going in the bad direction.
• and there are surely many other ways to help, I'm counting on you to be creative!

All data provided by this website is licenced under the Open Data Commons Open Database License (ODbL) also known as the "CC-BY-SA of databases".

The code of this website is licenced under the aGPLv3+.