codeforberlin / tickets Goto Github PK

Berlin has some amazing public libraries but the official website is quite dated and does not work well on mobile. This is why we started working on the opacapi project. It's goal is to provide a web-api that allows third party developers to get information about books and availability at the local libraries. One of the things that we could do with it is building a mobile app that allows people to check if their local library has a book that they are interested in.

If this sounds interesting to you you could start documenting what is still missing from the api for a first version of a mobile app to work. Feel free to just make comments here about what information one would need from the api and we will see if the api already provides this or if we still need to implement it. As soon as we have all data that we need we could of course also start working on the mobile app itself. 🚀

contactperson: @k-nut

There are a bunch of people who invest a lot of time into making public transport data more accessible. But there is a lack of PR so let's write a blog post. A good starting point https://github.com/derhuerst/vbb-modules.

contact person: @derhuerst

There is, e.g.,:

• https://www.frauen-gegen-gewalt.de/de/organisationen.html
• https://www.aktiv-gegen-digitale-gewalt.de/de/

but currently, those 'get help' are only sub-pages and not the focus of the websites. Come up with something that makes it as easy as possible to answer the question: "I need help, where can I go?"

Right now, there are several regional public-funded organization that monitor right-wing violence. Each of them uses their own way of reporting an incident on their website.

Here is a list of all organizations: https://verband-brg.de/index.php/ueber-uns/mitglieder

And here a visualization: https://vis.one/rechte-gewalt/

tasks:

1. define a standard format
2. scrape & convert all incidents

recommended techonology:

• https://morph.io/

contact person: @jfilter

Google collects data from our smartphones to derive their popular hours bar charts on, i.e., Google Maps.

Your task: scrape the data and build a database for all locations in Berlin (or Germany).

Other people also worked on it:

• https://github.com/apollojain/popular_times
• https://github.com/rak-trzustki/busy-hours

Test

Serious Machine Learning with Artificial Neural Networks: https://github.com/moovel/lab-opendatacam/

@derhuerst and @juliuste can tell the whole story.

The Bundesrat publishes its voting results on 16 different websites. Let's centralize them!

https://www.bundesrat.de/DE/plenum/abstimmung/abstimmung-node.html

So first, you have to extract the text from the text-based PDFs. Tools such as https://github.com/pdfminer/pdfminer.six could help. In the next step, you have to aggregate the results for each "Drucksache". Not sure how accurately the extracted will be. Maybe some manual data cleaning is necessary. To automacally scrape the results, you could put the scrapers on https://morph.io/.

Context: https://fragdenstaat.de/blog/2018/bundesrat-transparenz-klage/

Contact person: @jfilter

We sometimes tend to get lost in our own world when building tools.

What we have noticed in the past is that we are easily able to build tools that help people with less technical knowledge. An example for this could be a map in which a local NGO visualises some data that they have collected or building a small tool that helps a part of the local government do their job more easily.

I think there are many more examples where we could help. So if you know a NGO or have contact to the local government let us know and get us in contact. We would love to help people realise how they can use open data or how we can build civic tech tools to help them and the society as a whole.

contactperson: @k-nut

• Ideas how to make public transport more comfortable, more accessible, more transparent
• List of european long-distance transport operators, available API endpoints, GTFS feeds and client modules.
• list of libraries compatible with the Friendly Public Transport Format
• help crowd-source open public transport data
• a blog post about the VBB-hafas universe

We built an API to retrieve the trees in Berlin but we have never build something useful with it. Maybe you have an idea what to do with them? https://codefor.de/projekte/2016-07-18-be-street-trees-api.html

Contact person: @k-nut

On FragDenStaat are details about public bodies but not images. It would be cool to have for each public body an image.

You can get public body information here: https://fragdenstaat.de/api/v1/publicbody/

One idea, use the search on https://commons.wikimedia.org/wiki/Main_Page. Or use either https://www.wikidata.org/wiki/Wikidata:Main_Page or dbpedia-lookup to first get the item and then an image.

It would be cool to have an npm module, that gives me an image url to an public body name. Then it could be included in the frontend of FragDenStaat and also into the mobile app. But having a separate registry/database would also be fine. One could also think about somehow adding/merging all the public bodies into wikidata.

Contact person: @jfilter

The official „Nahverkehrsplan“ (public transport scheme?) for Berlin specifies that 80% of homes in Berlin need to be located within a given distance of a public transport stop where services run at least every 20 minutes (page 100 of the document, page 116 of the PDF).

The official scheme also cites a study to verify that the given goals have (not) been met in the past. We could either try to explicitly validate this information using census data, openstreetmap, public transport APIs or - which might make more sense IMHO - create a map that simply displays those locations within the city for which the requirements made by the „Nahverkehrsplan“ are not met.

There are many JS libraries and general tools available for working with the VBB GTFS data. But having additional (crowd-sourced) data will eventually help make public transportation in Berlin more accessible and convenient.

A few data sets we've started collecting:

It's unclear what projects are finished and what are still on going. To this, somebody should go over each of the projects at https://codefor.de/berlin/.

LobbyControl created a machine readable data about the political donations in Germany. https://lobbypedia.de/wiki/Parteispenden-Datenbank

It's common that politicians donate for their own party. So let's take a closer look into this. Go over each and filter out all the politican. But don't do it manually. Use some APIs either from https://www.abgeordnetenwatch.de/api or https://www.wikidata.org/wiki/Wikidata:Main_Page

related: https://github.com/jfilter/geocode-lobbycontrol-parteispendendatenbank

contact person: @jfilter

For the project openlegaldata.io we collected 10,000 judicial decisions (Urteile) from Germany's highest court the Bundesgerichtshof (BGH). The collected data is available over a RESTful API and bulk downloads can be also provided if needed. In the data you can most probably find some interesting stories which are worth researching. Are German judges more lenient after lunch time? Or how is the jurisdiction changing over the years? To discover these hidden gems we just need to start searching.

For instance, you could analyze what terms are most frequently used by the BGH judges:

But you probably can think about more exciting analyses. Be creative and feel free to talk to us about your ideas!

Summary:

• Data: http://openlegaldata.io/research/2019/02/19/court-decision-dataset.html
• Text analysis of judicial decisions
• Some legal knowledge useful but not required
• Data mining
• Visualizations
• RESTful API + Python SDK available
• Inspiration: Empirical SCOTUS https://empiricalscotus.com/

Contact person: @malteos

In Germany are a lot monuments but nobody really cares about the majority of them. So you make them more accessible.

To get the data: https://codingdavinci.de/daten/ -> search for "Landesdenkmalamt Berlin". Unfortunately, you can't link directly on it. In there they link to but I am not if there is more. https://daten.berlin.de/datensaetze/denkmalkarte-berlin-atom

There was something similarry done in Magdeburg but it's only a MVP: https://github.com/code-for-magdeburg/denkmal-magdeburg

contact person: @jfilter

https://offenerhaushalt.de/ is the German frontend for https://openspending.org/.

Sub Tasks:

1. find cities/regions that do not have an entry
2. formulate a freedom of information requests via the FragDenStaat API

https://fragdenstaat.de/api/v1/: With the API, you can also send requests. But it's rather undocumented right now. This will most likely change in near future

contact person: @jfilter

We should re-host our outdated, but well-known and/or well-developed, projects:

• https://github.com/stefanw/feld
• https://github.com/lieblingsplatz/lieblingsplatz

Es gibt scheinbar Daten unter https://ordnungsamt.berlin.de/frontend/dynamic/#!meldungAktuell

Aber laut https://lab.technologiestiftung-berlin.de/projects/datadive_cycling/index.html keine API und keine Karte.

Ein Projekt könnte sein, die Daten zu parsen und als API bereit zu stellen. Und sie dann zu visualisieren als Karte oder Auswertung.

This is about the German FOI: https://de.wikipedia.org/wiki/Informationsfreiheitsgesetz

There are only official statistics in Germany for the federal level (https://github.com/okfde/ifg-statistik). But there are none on state level (or maybe they are. So first do some investigation.)

But we have the exact number of requests for FragDenStaat: https://fragdenstaat.de/api/v1/

The task: Build a statistical model that estimates the number of foi requests (total, successful, failed) in Germany and per state/region/city.

@stefanw already did take a look into this and said that the numbers do not match at all. So maybe we have to clarify first what kind of data cleansing the administration is doing.

See also https://transparenzranking.de/ for regional disparities.

And here another visualization by FragDenStaat: https://arnese.github.io/IFG-Karte/karte5.html

Contact person: @jfilter

Berlin has set up a few bicycle detectors across the city. They also have a map, but it's not very intuitive, iFrame-based and doesn't really work on mobile devices.

We can do better. Let's make visualise the data in a meaningful way, on a mobile-friendly website. We may also create multiple forms of presentation for different purposes, or compare it to other cities.

Note: I'm not sure if the data is open. We may want to visualise is anyways (to show what would be possible if it was open) or talk to the administration.