Exploiting Open Data: the MELODIES conference
The MELODIES project concluded with a conference held in Brussels on the 11-13th October 2016. The event was called "Exploiting Open Data" and focussed on the creation and use of open data in the environmental domain, with a particular emphasis on the technical and entrepreneurial challenges and solutions that the MELODIES team and other in the field have faced.
Over 60 people attended the conference which was split into three parts: the first day focussed on the broad societal and policy aspects of open data use within the EU, the highlight of which was four keynote speeches from Barbara Ryan (GEO Secretariat), Ed Parsons (Google), Emma Truswell (ODI) and Luis Daniel Ibáñez (ODINE) and a following panel discussion. The second day gave a platform to the developers of all eight MELODIES services as well as the three tool/platform providers to summarise and contextualise their work in the broader context of the open data landscape – a key aspect of this day was the demo session time allocated in the afternoon for one-to-one discussions between MELODIES partners and interested external attendees. The final day saw a series of tutorial training sessions on the platform and tools provided to the MELODIES project, this allowed both internal and external potential users to have a guided tour of the technical resources developed by the MELODIES project.
There were many discussions at a variety of levels throughout both publicly and one-to-one throughout the three days. The key themes and messages which emerged were:
- The community is excited about the use of open data to drive new products and services.
- Many challenges remain concerning the usability of open data. More needs to be done to make data easier to find, access, understand and use. However, it is not clear whether this task should be done by the original data provider (who would need more resource to do so) or value-adding third parties (who need a viable business model to justify their costs).
- A key issue, particularly for private-sector usage of open data, is the question of quality, both of the data and of the service.
- Linked data is an area of research and work that has considerable momentum, but its complexity appears to be deterring potential users at this stage. One strategy that some companies are adopting is to use Linked Data as a tool primarily for system developers (taking advantage of the highly consistent and structured nature of the data), whilst providing end users with simpler APIs and more familiar formats.
The discussions on Day 1 focussed on the opportunities that Open Data can provide to entrepreneurs and – by extension – to the population at large. These are summarised here and illustrated with photos and tweets published on the day. You can see all the tweets from the event in this Twitter Moment: https://twitter.com/i/moments/788777053130817536
José Miguel Rubio-Iglesias: Welcome from the European Commission (image credit: @terradue)
Jon Blower: Overview of the MELODIES project (image credit: @OpenAccessEC)
Hervé Caumont: Engaging Earth observation in the Platform economy (image credit: @EmmaTruswell)
There were introductory talks by the MELODIES project officer José Miguel Rubio-Iglesias, project coordinator, Jon Blower and cloud services provider Hervé Caumont which together set the scene of the MELODIES project and the political/technological landscape in which it has operated.
Barbara Ryan: The Group on Earth Observations (GEO) and the Importance of Broad, Open Data Policies (image credit: @MelodiesProject)
The keynote talks began with the GEO Secretariat Director Barbara Ryan, who described the enormous economic benefits available to governments and public bodies of simply opening up data, using the example of opening up Landsat data in 2008. She also described the work of GEO now in realising that same vision for all Earth observation data, remotely sensed or otherwise.
Ed Parsons: Open Data - A Hi-Fi approach (image credit: @MelodiesProject)
Ed Parsons from Google spoke next about the importance of developing technological frameworks which support the easy use of open data. He used the example of the amplifier in a hi-fi system, the music from the tape deck is unable to be heard without the amplifier, Ed called on the community to generate ‘amplifiers’ for open data, allowing it to be accessed and used to its full potential.
Emma Trusswell: Open Data, Innovation and Public-Private Partnerships (image credit: @MelodiesProject)
Emma Trusswell spoke about the work of the open data institute in engaging local communities with the opportunities of open data and helping them over the barriers to use. She explained that many of these barriers can simply be down to language (‘the word “data” sounds complicated’) or lack of information about the datasets available. She used an ODI project fostering new open data start-ups in Mexico City.
Luis Daniel Ibàñez: ODINE: Creating an European Ecosystem of Open Data SMEs (image credit: @terradue)
Luis Daniel Ibàñez spoke about the ODINE project, an H2020 project which supports the development of open data businesses by awarding grants to successful applicants. Luis gave some examples of innovative project supported by Odine, including the creation of public benches in European cities which use moss to clean up pollution in the air, this moss is kept alive by smart, open data technologies.
All of these discussions were summarised in a piece of artwork called ‘visual minutes’. This hand-drawn poster was drawn as the speakers were talking and once complete it remained in the hall for the duration of the conference.
Visual minutes in progress on day 1 (top) and completed (bottom)
Day 1 concluded with a panel discussion coordinated by Jon Blower and joined by all the other speakers from the day. Questions and answers ranged from topics from the use-ability of linked open data as the technologies stand today to how the community can build itself up around these open data sources and drawing in new users.
On Day 2, each of the MELODIES service-development partners gave 20-minute presentations on their service, the benefits that open data has afforded it and future developments. There was a lot of audience interaction with questions following nearly all presentations. After lunch, representatives from Terradue, the University of Athens and the University of Reading gave brief 10-minute talks on the technologies that they have developed. All of the presentations are available online as videos and the presentations are also available on the MELODIES website.
Snapshots from Day 2 of the conference. Line by line, from top left to bottom right these images show, Gerardo López Saldaña, David Leaver, Tomas Soukup, Marion Buddeberg, Ugo di Giammateo, Ricardo Goncalves, Gunter Zeug (left) and Conrad Bielski (right), Charles Calkoen, Francisco Batlle, Cesare Rossi, Manolis Koubarakis and Jon Blower. Images taken from the twitter feeds of @MelodiesProject and @terradue
The afternoon of day 2 was spent in demo sessions where each of the partners who had spoken during the day had a stall with a professionally-designed poster and flyers. Conference attendees were encouraged to talk to the MELODIES partners about their work and about potential business opportunities. All asked partners reported that this demo/networking time was very valuable to them and many had collaboration/business opportunities to follow up.
On the final day, attendees were given an opportunity to learn about the tools developed in the MELODIES project through six hands-on tutorials. Joining instructions (e.g. which software to download and install beforehand) were provided to all attendees well in advance.
The first two tutorials were held by Cesare Rossi and Hervé Caumont and showed users how to access the Terradue cloud, set up and perform some processing. Contact Herve Caumont for more information.
The third tutorial was conducted by Dimitrianos Savva and Georgios Stamoulis from the University of Athens on using GeoTriples and Sextant to create linked data from a postgres database and view it using sextant. The fourth tutorial was run by Konstantina Bereta on Ontop Spatial, an extension of Ontop which is designed to allow users to work with raster data as though it were linked data.
The fifth session was run by Guy Griffiths on running and using ncWMS to display NetCDF data in a browser. The final session was conducted by Maik Riechert on CovJSON – a cutting-edge approach to displaying and manipulating coverage (raster) data in a browser.
Maik Riechert leading the covJSON tutorial (image credit: @MelodiesProject)