Dublin2013

Dublin 2013
As part of International Open Data Day, a Dublin Hackathon is being organised by the Open Data community - supported by Dublinked and Fingal Open Data

We'll add more details here as planning progresses - also, keep an eye on the Open Data Ireland Google Group for updates.

Location
The event will take place in the Wood Quay Venue in the Dublin City Council Civic Offices.

Location on Google Maps

Location on OpenStreetMap

Time
We'll be starting on Saturday, February 23, 2013 at 10:00am and finishing at 5:00pm.

Format
The following approach to the event has been proposed on the Open Data Ireland Google Group -

People with ideas make very brief outlines of their ideas by giving a quick 2 minute pitch (project pitch guide) at the start of the event to see who else is keen to work on the idea and perhaps jog some people in the room that can acquire the data needed.

Facilitate the following types of people to work together on projects 1) the people with the data problem 2) the people currently holding the data 3) the people who can turn the data into some sensible representation / interpretation of reality. (provide 3 different colours of stickers - one for each type of person to wear, to facilitate forming of groups)

{| class="wikitable" border="1"

Participants
Register to attend at https://tito.io/open-data-ireland/international-open-data-day-dublin-hackathon-2013

People
Add your name here if you are attending ...


 * Dublinked
 * ActiveCitizen
 * Richard Cantwell
 * Denis Parfenov
 * Sara Burke
 * John Muldoon
 * Robin Cafolla

Projects
If you have a project you want to work on, please add it here.

Higher education statistics for Republic of Ireland.
They are just dumped in a collection of spreadsheets here: http://www.hea.ie/statistics. Any help improving the presentation and making the numbers more attractive/meaninful/fun would be greatly appreciated.

Projects to provide solutions for Local Communities
e.g.
 * Parks
 * Census
 * Libraries
 * Crime
 * Politics
 * Local Public Expenditure
 * etc.

Health Stats
Health Stats web application/data visualisation [inspired by NHS hack day http://www.nhshackday.com]

There is a HSE website at http://www.hse.ie/eng/staff/Healthstat/about/ that is meant to track the performance of the various hospitals but to honest I can’t really find much that is useful on it – why don’t we try to rebuild it as a free, informative, useful tool that anyone from patients to doctors to citizens to journalists can use to easily access information on hospital performance?

Step 1. - Acquiring the data: We would need to have access to the raw data that is used to produce the pdf that is published monthly on site – which currently gives a meaningless traffic light type indication of how a hospital is doing. But the data underneath could be used to build something really good.

Step 2. – Parsing the data: Once we have the data (!), there will be some effort in cleaning and exploring the data to see what exactly it affords with regard to what information we can get from it and what form an application built on it would take. We would need to ask ourselves what users would want from this as well.

Step 3. – Building an application: I’ve looked at bits and bobs built with the http://d3js.org/ JavaScript library and they look really great. But others here might have alternative suggestions. We could setup a github and get the coders in the group and of course anyone else who wants to, collaborating and building this thing.

Map Kathmandu
International Open Data Hack Day event being organised by MapBox: http://mapbox.com/blog/open-data-day-mapping-kathmandu

This might be of interest for people working in disaster management / emergency planning. The plan is to improve the (fully open and reusable) OpenStreetMap map of Kathmandu in Nepal from Satellite / Aerial imagery. The event is being hosted in Washington DC but remote participants are encouraged. Richard Cantwell is familiar with the project and can provide local assistance.

Alternatively we could look at improving the OSM data for Ireland, perhaps adding medical facilities or other items of interest which might be currently missing from the map. Here is an example of the OSM map with Hospitals (blue) Doctors (green) and Clinics (red) highlighted:

Open Food Facts
[] - "we would love to have more Irish food products :-) The French version is going very well, but we need help to get English to take off"

The Museum without walls
Geochronal mapping of historical data-objects from Dublin City Public Libraries, National Archives, etc on a Leaflet/OpenStreetMap base, to provide affordance for collective remembrance through wikihistory, and area curation and destination development for tourism.

Time-series map imports (hopefully from MyCity - DCC and city council maps) to show morphogenetic change in urban environment, historical development, and psychogeography.

'Hard' data from providers, 'soft' data / anecdata / user submitted material from user-base.

Today's hacking tries to create a foundation for the capture of input from site users. The ultimate goal is to allow users to click at a location on the map and to infer an area about which the user will then contribute content. In this implementation, click locations will be matched against named buildings present in the OSM data set.

Completed: Ability to look up the clicked co-ordinate and identify, by name, any building found.

To do: use the geometry returned by the lookup to display the building outline. Popup panel presenting meta data for the building with option for user to add content in a manner yet to be agreed with the users.

Open Data Finder
Request and search for open data + repository for data ans links to data. Outreach to the general public to find what data people actually want or need.

Hospital Dashboard – Health Data
An online informational “dashboard” that includes
 * rates for infection
 * readmission (what we call “bounce back”),
 * surgical complications and “never event” errors (mistakes that should never occur, like leaving a surgical sponge inside a patient)
 * hospital’s annual volume for each type of surgery that it performs (including the percentage done in a minimally invasive way)
 * patient satisfaction scores
 * private / public patient waiting list by hospital / procedure

Crystal (Swing States)
A web application to highlight inherent democratic deficits by showing voters in Ireland the constituencies and seats that change or are likely to change in an election.

The app will present historical election data and analysis of voting patterns to show voting instability and the regions where a small number of individuals can make a difference.

Conversely it will also show how through voter trends it is unlikely that your vote for a particular candidate will influence an election.

Process
The team split up the jobs into


 * Data Collection
 * Data Provision
 * Data Storage
 * Design

Over the course of the day a prototype application was built. In the process the 2011 Irish general election data was scraped from Wikipedia and is found here:

Data (JSON)

This data is released under the Creative Commons Licence.

Out Come
The demo application color-codes the winning parties in each constituency based on their first preference votes:

Application

This is envisaged as the start of a larger project to finish the app and allow users to see where their vote counts and where it does not.

Information
For more information or to contribute please email [mailto:robin@zombiemongoose.com Robin]

Which areas of the city are furthest from open space
This project was carried out by Richard Cantwell

I worked with OpenStreetMap data for Dublin City to discover areas that are furthest removed from leisure facilities.

This was a multi stage process using online tools and Commercial software. The entire process could be done with Open Source software but as time was short I worked with the tools which I am most familiar with.

The first step was to get may hands on the data. This was easy enough. OpenStreetMap has a number of ways of providing data, direct download or via an api being the most common. For this project I went with the api option.

There is a visual query editor for the OpenStreetMap API (OverPass) called OverPass Turbo which generates data in geojson format. Downloading this data and importing it into my GIS weapon of choice (MapInfo) gave me the locations and shapes of areas tagged with 'Leisure' from the OpenStreetMap database.

(Note re: the maps below, they're very quick and dirty and also quite large, about 1.5Mb each)

A quick output map showing this can be seen here.

The next step was to build a grid, I chose a 250m grid to give a reasonable amount of detail while keeping the processing time sane. Heres a quick map of the Grid

I then calculated the distance from the centre of each grid cell to the nearest leisure using an in-built function in the software. Here is the result.

For fun I then did the same process for pubs, here's the distance grid.

I then did a quick calculation to see which parts of Dublin are closer to a pub than any kind of leisure. The results, for a city renowned for it's drinking, were a little bit surprising.

'''Key issues ''' Obviously this analysis is entirely dependent on the source data. OpenStreetMap is a wonderful resource, but it isn't perfect - no data is. But anybody can edit the map and improve it's accuracy. To find out more have a look at the beginners guide.

Also I only had time to generate quick .jpg outputs. Given more time a proper web based interactive map could be built.