Experienced limitations of big data

• A major limitation discovered with open data is its lack of detail and relevance (usability) for the particular service contexts. While open data may reveal some strengths and weaknesses in particular service areas and geographies it still often is too generic to be meaningfully applied in improving and tailoring services for the great diversity of citizens as recipients of the same services.  For example, the Estonian start-up and creative development scene found challenging to provide developers with real-time, reliable and well-presented public data.
• CoSIE evidences that it has been more meaningful for service actors to hear individual lived experiences, “seeing a person behind the voice” or extracting knowledge and insights in extensive thematic dialogues, workshops (Social Hackathons) or focus groups with the help of neutral facilitators.
• Contrary to its intention, open data is not immediately accessible to those that are digitally excluded or those that may not possess required technical and analytical capabilities. That sustains disempowerment (Jamiesson et al., 2019; Jallonen & Hello, 2020) or else requires extra resources and intermediaries. In CoSIE, public service officials rather than users could access and were more eager to explore open data.
• A major threat with too heavy reliance on large sets of open data is that such data is not always created intentionally but is increasingly generated automatically based on algorithms and user digital footsteps.  Too heavy reliance on it by service providers might dehumanise the user profile and, in the end, the services, by making user a passive subject.Data in not neutral, it is interpretable and the user as well as service provider should have a say in its interpretation. Besides, using open data for innovations may require resolving heterogeneity of interests and eventual conflicts (Jallonen & Hello, 2020).  (CoSIE examples?)
• One of the main obstacles to publish open data is the fear that the privacy of individuals will be violated. Data anonymization is a type of information sanitization whose intent is privacy protection.  In Popowice pilot, Poland, the open data application initiative aims to help with collecting information about the problems of the residents of Popowice and what they would like to change in their housing estate. The sensitive personal information such as name, surname and telephone or contact e-mail needs to be suppressed from open datasets. Yet sometimes data anonymization is not easy especially when users are a small group with individual profiles and in disability services in the Swedish pilot. Complexities with anonymization and legal barriers prevent data sharing even between interlinked agencies
• To be useful, open data requires transparency andwell-defined procedures from gathering to publishing. This is easier said than done as data can be owned or created by different organizations, which typically complicates the processes. Sometimes there is a lack of insight in how the processes could be modified to improve them. In addition, accidentally published private and sensitive data probably damage the organization’s reputation and may even lead to legal liability.

Reflective exercise:

Open data and service or social innovations: Is it so that the “plausibility of open data is only to be realised at a community level – that is where the interest and abilities have shared beliefs and intentionalities”? (See Jamiesson et al., 2019, p 366)

For inspiration – take a look at the Spanish case.