Develop dataset criteria

The problem

Throughout the implementation of an open data plan, public sector actors will need to understand what can be considered “high-value datasets” so that publication of these datasets can be prioritized. This best practice sets out a number of criteria that can be used to determine what are ‘high-value datasets,’ taking into consideration the likely re-use of open data and to help governments understand which datasets to prioritise for publication. The characteristics of ‘high-value datasets’ are seen from three perspectives: re-usability, value for data owners, value for re-users. Therefore on-going engagement with the greater open data ecosystem is encouraged.

The solution

Specify Methods Of Determining The Prioritization Of Data Release, Open Data Policy Guidelines, Sunlight Foundation: This resource to aid governments in creating an open data policy, recommends putting forth a process to determine which datasets to prioritize in their release. The resource also includes information on examples of such processes from U.S. cities to draw upon in implementation.

Build On The Values, Goals And Mission Of The Community And Government, Open Data Policy Guidelines, Sunlight Foundation: Goal setting in open data programs can guide the prioritization of dataset release, how to release and how often. Goal setting should be the basis of program implementation.

Methodology, Global Open Data Index 2015: Open Knowledge International (OKI) process for determining the methodology for the Global Open Data Index engages civil society and sector experts to identify which datasets are crucial to the sector, how often and in what aggregation levels. The methodology is updated annually and user-driven from the civil society actors who will ultimately be data consumers. This process could serve as model for government piroritization or the public sector could consider the datasets recommended by Index as it determines which to release and how. The Global Open Data Index utilizes an online forum to facilitate discussion between data users and providers. This forum could be used or a similar one could be emulated by public servants implementing this best practice.

Draft.dc.org: The city of Washington, D.C. put its Open Data Policy online for comment through madison.io, a platform developed by OpenGov Foundation. This initiative exemplifies an innovative approach to getting all stakeholders at the table and could be a potential way to implement this best practice.

How to open data, Open Data Handbook: Open Knowledge International’s resource outlines a process to gather public and CSO feedback on which datasets should be prioritized. Those that are in demand by these data consumers help inform what datasets are “high value.”

School of Data: This global network of data literacy practitioners, could be drawn upon and partnered with by public sector actors implementing open data programs to understand what types of high value datasets are in demand by civil society.

OpenSpending: The Share PSI best practices identifies public spending information as being a high value dataset based on Share PSI workshops. The public sector could engage with the OpenSpending platform and community in the implementation of the Share PSI directive.

Open Data Dataset Starter Kit, Center for Government Excellence: This guide, developed by the Center for Government Excellence at John Hopkins University, helps governments rapidly identify high value datasets for publication.