Open Data network: the Challenge of Interconnectivity

Nicolas Terpolilli
5 min readDec 7, 2015

--

The fundamental model of the digital economy is a network. As shown by the architecture of the Internet itself, whose reliability and robustness grows with more and more nodes, we also have Facebook, a network whose interest increases for a user according to the number of his or her friends present.

The Network: the Heart of the Digital World

It is for this reason that we are now allowing our users to search for a dataset anywhere on the OpenDataSoft network. We are now offering access to a sort of dataset library, collecting the datasets of all of our users in one place. This means that it is now possible to retrieve any dataset from one place, produced by our clients: all of them, without duplicating, without losing metadata, and without risking security. Not only that, this is all possible while maintaining the possibility to filter parts of the data.

This prospect facilitates even further reuse of the data of a given producer, which is generally one of the objectives of any Open Data initiative. It makes it even easier for anyone — especially with a free OpenDataSoft account — to take advantage of mapping tools, as well as to create graphs and widgets to help discover data. Once again, OpenDataSoft speeds up the process between the publishing of a dataset and the creation of services based on these data, by anyone, anywhere.

This network of datasets leaves new and exciting models of data distribution all to the imagination.

MaVille: automatically generated portals made possible by the network

After the passage of the NOTRe Law (article in French), a large number of French cities will begin the process of publishing their data. Therefore, we have decided to begin this process by automatically generating portals for cities with more than 20,000 inhabitants.

To do this, we searched on data.opendatasoft.com for data at the local level. From demographics to 4G coverage, and even job offers, plenty of useful datasets can be made immediately available to citizens, without them even needing to go search within a dataset to find relevant information.

All we had to do for each portal was to add a dataset from the OpenDataSoft network, filter by the commune’s INSEE code, and create a homepage that automatically picks up important information. This is all possible thanks to a few copy-pastes with our widgets!

Now, let’s imagine that:

  • Large industrial groups are just beginning to take part in this initiative,
  • Any citizen can be inspired to go create their own dashboard,
  • Different levels of administration have a certain coherence in their respective Open Data initiatives,
  • Comparing and thus emulating different organizations is rapid, and doesn’t require hiring a technician;
  • All of this is done without losing information when data is updated, for the simple reason that there is one single dataset that serves all of these uses,

Then, we are starting to see the effects of an Open Data Network and all of the potential synergies in the world of data.

Open Data Network: The Global Rise of Open Data

Open Data is advancing, and rather quickly at that. One must only look at the first portals, which offered only lists of downloadable links — Amazon was once but a simple list of books in HTML. With the multiplication of open data and their reuses, solutions are evolving. Gains in efficiency are happening at every level of the value-chain.

Attempts at researching the effects of networks in the data are far from being a recent phenomenon. The movement of linked data and the semantic web are the best examples of this. If we think that this vision is very strong and carries a lot of interest, two things could be what slow all of this down:

  1. Cleaning and preparing linked data is still a job for experts in the majority of cases, and requires a lot of time. This will evolve with the multiplication of tools and with clear definitions of ontologies, but it is still a process that cannot be industrialized.
  2. Interrogating and analyzing data is still a major challenge for many people. The language of the SPARQL request is very powerful and allows for the appropriation of the data, including its structure. However, SPARQL requires training, and the tools that would allow for everyone to take advantage of it do not yet exist.

The semantic web is clearly an inspiration for all of us, and this Open Data network is a first step. In the very near future, OpenDataSoft will allow for the linking of data. Open datasets from the network will soon come and feed the web of data.

OpenDataSoft has always wanted to make data, their spread and their reuse accessible to the largest possible number of people, multiplying efforts to supply tools accessible that do not require a lot of training. Facilitating even further access to data, OpenDataSoft allows for the advance of these three aspects all at the same time.

Originally published at www.opendatasoft.com on December 7, 2015.

--

--