Towards a User-driven Open Data Ecosystem?

During the ODECO training week in Zaragoza in March 2023, whose focus was on developing user-driven open data ecosystems, we were reminded of a conversation we had many years ago with the CEO of a location-based marketing company with regard to location data. Some questions were raised. How was location data obtained? For what purposes? And especially, with which parties was the data being shared? Is there anyone who keeps track of the journey of the data? Where does “our” location data that we share with, for example, our mobile provider go? With which parties is it shared and reshared? At that time, we felt quite unsettling that nobody seems to know these answers. Since then, the traceability of our location data (and personal data in general) remains at the back of our minds. What could be a potential way to preserve the path of the data journey? Would a type of watermarking be a way? Or is it installing a data tracker in each of “our” data that sends automatic notification each time “our” data is used or is shared with another user?

Although the questions above are about personal data, hence are not open, we found those questions can also be applied to open data, that is how to keep track of the use(r) of open data. Open data can be defined as data that can be (re)used without any restrictions at the point of use. Key for successful open data ecosystems is that data providers and users are in contact with each other. However, one of the shortcomings of the open data implementation is that data providers lose contact with the users of their previously non-open data. Some open data providers address this issue by requiring users to register before enabling them to use the open data.

However, by definition of open data, registration is considered as a restriction to the use. Nevertheless, the practice of certain data providers such as by the Danish Agency for Data Supply and Infrastructure shows that registration is not a significant barrier to data use for many users. There are also other approaches to identify open data users. For example by organising user events, creating user groups (e.g., http://odug.org.uk/) and forum (e.g., https://opendata-forum.cern.ch/), carrying out surveys (e.g., https://library.wur.nl/WebQuery/wurpubs/reports/440546), and implementing helpdesk services (e.g., https://geoforum.nl/).

In Zaragoza another possible approach was also brought to the table, namely share the open data through APIs that belong to a data space. In this data space, participants may add to standard open licenses the right to be notified when their data is reused through an API. That is, other APIs that build on the open data should first adhere to the rules of the data space and, hence, notify the upstream open data API of the reuse of the downloaded open data. Once the network of reuse notifications is built, it will be easy to peer-to-peer request to the APIs to determine the reuses of a specific dataset across the network, making the chain of dataset reuse visible to the user(s) who contributed their data.

As one of the sustainable open data ecosystem’s cornerstones is the shift from
conventionally supplier-driven to more user-driven open data ecosystems, the challenge of engaging users of open data in the ecosystem is very often brought to the forefront in the ODECO consortium discussion and research.