Our research investigated how to motivate big data owners to share their data and how to get external stakeholders interested to exploit it through a case study. The purpose was to identify alternative solutions to open the currently closed invention database owned by a Foundation. We utilised semi-structured interviews with key stakeholders as a primary data collection method. As key results we found interested parties for the database, and drafted possible business models.
The first stage was getting familiar with the case through reading database documents and web sites, discussing with one of the owners and trying the database in use. The second stage consisted of 13 interviews with database owners, possible users and open data experts. This was to understand the owner view of opening the database and the related barriers, the user needs and wants and possible solutions to overcome barriers. Based on this knowledge, we developed alternative solutions for opening the database. Then, we collected feedback from the owners and possible users for these solutions. Finally, based on the feedback received, we formed sustainable business models for this case.
In this case study, we relied on semi-structured interviews that were conducted either by phone or face-to-face. These interviews always involved the project member responsible for the case study and an interviewee, who was either a database owner, possible user, open data expert or a representative of a related database. The biggest bottleneck was to find the time for the interviews, as the contributors were busy. In several cases, the initial interview needed postponing. A success factor of the interviews was the good motivation of interviewees to participate in the study. The owner was interested to get financial benefit from opening the database and the possible users valued the database content and were interested to learn more about it.
The key success factor was to initiate an open dialogue with the interested parties for identifying and developing alternative solutions to an open database. The key barriers for opening the database were mostly technical, IPR and privacy issues. For instance, the database does not enable creating different types of user rights, such as read-only. Implementing this feature would be costly. Moreover, as this database contains confidential information, it can only be shared with researchers, for research purposes, and they will also need to sign a non-disclosure agreement (NDA) to access the database.
Based on our interviews with open data experts, researchers often think that open data means the same as public data, thus seriously limiting the motivation for researchers to open their research data. When the data is highly confidential, as in this case, there are, possibilities of opening data for users for a restricted purpose by signing NDAs. The owner needs to define the access criteria, and can also put access fees for usage. Aside from the financial beneftits of sharing their data, the researcher could also gain from additional publications, research citations and possible new partners to collaborate with joint research projects. Some researchers are also reluctant to share the data as it means having to input the database content and defining metadata for others to find it. There can also be technical barriers for sharing the data as in this case. However, these can usually be all solved as long as the database content is valuable enough that it pays off to do this work.
- Inform researchers about various possibilities to open data – open is not the same as public and train them to implement this into their practice.
- Motivate them to open data when the database contents are valuable – there are both financial returns and academic merits to be achieved.
- Rush into opening data – it pays off to carefully evaluate alternative solutions for opening research data – weighing pro’s and con’s of different models. In order to do this, always identify the needs and wants of potential user first and clarify the barriers for opening research data.