We need better Open Data Interfaces

By pietercolpaert

2017, October 10

Better?! Come on, you must be more precise with your feedback!

A scientist

The technology to publish data as broad as possible – Open Data – is slowly changing the world. For example, it’s changing how we make decisions from gut-feeling to evidence-based. It also makes our lives easier as route planning applications can consider more transport options than our brain, connected to the Web taking into account the latest real-time updates. It is also the start of new business models, exploiting the growing mass of data that can be discovered online.

Every dataset, how small it may be, may contribute at some point to finding an answer to some question. Yet, we have a big problem. Almost nobody* is reusing the small datasets published on Open Data portals. At least, if the data would indeed be useful, we should be able to notice evidence of data reuse in our daily applications. Is this a huge fallacy for those who claim there is economic and societal value in their freshly published datasets?

For the years to come, Open Data researchers will need to prove that publishing data publicly is in the interest of both existing and new business models. I accredit the lack of adoption today to a simple cost-benefit analysis: would you reuse a dataset when it costs you more to integrate it in your system than the benefits it would give to your end-users? I know I wouldn’t.

Now that I am a postdoctoral researcher, I am in charge of leading a small research team. For the next years, our new “Linked Open Data Interfaces” (LODI) team is going to research how to lower the cost for adoption. I have already found funding to expand this team with junior researcher, and I am now looking for information/software engineers that would like to join.

* To be fair, there is some uptake, but not to the extent that data gets picked up automatically by any interface that could benefit from it. How can we advance the state of the art in data publishing, so Open Data really thrives?

What we do

We have a vision for a better world, where knowledge creates power for the many, not the few. If anyone can access public data, rather than a few with the deepest pockets, we create a more level playing field for building on top of the single and largest human knowledge base. You may recognize the slogan from the NGO Open Knowledge, with whom we work together in projects such as OASIS and open Summer of code.

We research how to create better Open Data interfaces. We define “better” as how much easier it makes the lives of reusers. Our job entails making sure the interfaces are more reliable (and thus more cost-efficient for the data publisher), interoperable, queryable and discoverable.

You can compare our Linked Open Data Interfaces team with a university hospital where a doctor does specialized consultation on the one hand, and ground breaking research on the other. On the one hand, our inspiration for research comes from real experiences in the field. We work with, among others, the Flemish government on projects to decide on the new open information architectures. In bilateral projects, organizations can hire our team for short projects to publish data in better ways. But we are also involved in longer projects, where similar data interfaces need to be studied.

More information on such projects (in Dutch): data.vlaanderen.be, Smart Flanders and Local Decisions as Linked Open Data

On the other hand, we do fundamental research. For example, with Linked Data Fragments we introduced a way to study the trade-offs on Linked Data Interfaces, we study what features of new technologies such as http/2 could be exploited for data publishing, or we are studying how to build a crawler that would be able to plan travels by public transit options world-wide.

More information: the website of Linked Data Fragments and the website of Linked Connections.

Apply for a job at our lab

You are joining team lodi (Linked Open Data Interfaces) at the Internet and Data Lab in Ghent: a team of 5 researchers that want to see a world where knowledge creates power for the many, not the few. Your ambition is to become one of the main data architects within Flanders.

The first deadline is to send an e-mail to Pieter Colpaert with your CV and motivation letter before the 1st of September 2018. We envision the first hire to start in October, but earlier or later is possible.

Other practicalities...

  1. You have a masters degree and are experienced in computer science or engineering.
  2. You are living in or willing to move to Belgium
  3. You speak Dutch and live in Flanders. This is considered a big advantage, as many projects will be in collaboration with the Flemish government. But do apply if you are able to speak English as well! We will have other projects as well, and we are hiring 3 people in total.
  4. The salary is fixed and depends on the years you have worked so far. We can make an estimation of a net salary if you contact us by mail.
  5. The research you do can lead to a PhD (although this is not a requirement).

Things to read

We cannot wait to hear from you. If you however want to read up on our work, here’s a list of relevant links:

An introduction to why Open Data is important, from the Kardashev scale to tracking seagull Eric, can be found in one of my recent talks published on youtube.