We are very excited to announce the next “Open Food Data meets Smart Kitchen” Hackdays on the 7th - 9th of September in Zug! Are you eager to develop new projects to make the food industry more transparent, innovative & sustainable? We are looking for bright minds from all walks of life to brainstorm, discuss and collaborate on innovative solutions. Join this prototyping adventure via https://food.opendata.ch/#hackdays.
Learn to work with open data and help to prepare resources for the upcoming Open Food Data x Smart Kitchen Hackdays.
We encourage people of all skill levels to take part in this event facilitated by the School of Data. The workshop will be followed by opendatabeer.ch from 18:30. Whether you are a data connaisseur or beginner, this pre-event is the perfect chance to get ready for a hackathon!
The workshop is happening today!
If you have registered, you should have received instructions about the location - and if you can’t make it in person, you are still welcome to join our team chat and follow along virtually. Instructions can be found here on our website.
A short summary of the workshop follows, and there are additional notes on our team channel “Expedition”, which you can join here. If you have any questions, please feel free to reach out to us here, there, or via food at opendata.ch
Kick off: Character Sheet
Each of us began by completing a School of Data character sheet to present ourselves, our technical skills and break the ice.
Through this we quickly started discussing the mission of Opendata.ch, Rufus Pollock’s new book, the possibility of making predictions with data (e.g. A Cautionary Tail: A Framework and Case Study for Testing Predictive Model Validity, Peter Casey et al 2018) as well as the pros and cons of data tools like R.
@oleg pointed out: Today’s goal is open data activism. We want to create value for the greater good with your skills. “Data is a tool for understanding the world.”
Introduction: Open Food Data Program
You can find the slides that were presented here:
- Our goal is to make the food system more #transparent, #sustainable and #convenient
- Let’s save the big challenges of our food system together:
- The United Nations (FAO) estimates that about 815 m people were suffering from chronic undernourishment in 2016.
- The World Health Organisation estimated in 2017 that 2.1 billion people lack safe drinking water at home.
- The Swiss Federal Office for the Environment estimates that about a third of all edible food in Switzerland goes to waste.
- PwC Study estimates that we’ll need 35% more food, 50% more energy and 40% more water until 2030.
- Our 3 key activities are:
- We are trying to improve the open food data program and want your feedback on how to do it.
- Our incubation selection criteria are:
- be innovative
- have a positive impact
- stay feasible
- strengthen open data
What is Open Food Data?
How open food data should be described in the best case:
- Issue Date
- Modified Date
- Temporal Coverage
A few helpful links to learn about open data and how to work with it:
- https://frictionlessdata.io/specs/ describes how open data should be made available and described
- CKAN is a management system for the storage and distribution of open data.
- http://goodtables.io helps to validate data
- https://github.com collaboration tool to work with software and data
- https://datahub.io collaboration tool similar to GitHub with focus on data
The goal of this workshop is to prepare new food datasets to make them open & easy to use. We call prepared datasets: “data packages”.
You can find a general explanation of the data packaging process at schoolofdata.ch:
First, we all need to research and select a dataset that meets our goals and criteria. Then we go through this general process:
- Figuring out if the data is open & legal to use (license)
- Downloading, checking, aggregating, cleaning, and exporting the data in open formats
- Creating a “Data Package”
a) Write a README
b) Prepare a
- Share the results via GitHub / DataHub.io / Datacentral
Note the tools being made available for step (3) from DataHub.io and our “boilerplate” data package, which you can fork and edit right in your browser with a GitHub account:
In this manner we started packaging weather data for agriculture (Agrometeo), which you can now find in progress here:
Discussion: how to present open food data?
Questions that have been raised are:
- Would you be able to work with a data description on GitHub or DataHub or Datacentral? - how else would you like to see data presented?
- The storytelling & context of a dataset is important - how can we get people excited? → Add a section to the data “why this is important”. But e.g. for the Hackathon people have a specific target or user case in mind. In this case it’s not important to have a “why this is important”.
- How to make it easy to suggest new datasets?
- How does one know if a dataset is current? As soon as one becomes a data user - one is interested in the maintenance of a dataset. Create a relationship with the data owner / provider.
- What is the ideal data format for open data? --> an “open format” --> consult the Open Data Handbook, check out the Open Data Index, and look up the new Frictionless Data standards:
As a next step, with the insights from the workshop, we’ll discuss how we want to plan the future of the open food datasets.
- Make more prominent on food.opendata.ch? → Navigation “Open Food Data”
- FOOD DATASETS: Organise by topic? Or just have search function? Or sort by?
- ADD FOOD DATA: Make it easier to add dataset suggestions + link to data packaging explanation
- FOOD DATA HISTORY: Track growth of data availability?
- A good example: https://codingdavinci.de/daten/
Superb - thank you very much for the notes @Nikki
Kudos to everyone who came out for the Open Data community event that follows. Hope to see you in Bern again soon. Next station --> ZUG !
Additional impressions via Twitter.