This chapter contains the following key action.
1. The Government’s default position is that departments and public bodies datasets must be made available to the public.
3.1 Definition of datasets
The Policy has adopted a broad definition of ‘data’. The definition for data refers to datasets and databases owned and held by Victorian departments and public bodies and stored in formats including hardcopy, electronic (digital), audio, video, image, graphical, cartographic, physical sample, textual, geospatial or numerical form.
The Policy mandates that datasets must be made available in machine readable, reusable and open formats. The Policy also applies to data made available in the form of an Application Programming Interface (API), web service or data tool (as long as the tool has a machine readable output).
The definition of datasets is intentionally broad to ensure all agencies consider a broad range of datasets to be made available.
Agencies are encouraged to review existing and new datasets and determine whether it is appropriate that they be made available.
3.2 What must be made available under the Policy
The Government’s default position is that public sector datasets must be made available to the public, unless access is restricted for reasons of privacy, public safety, security and law enforcement, public health, and compliance with the law.
Agencies will be required to implement the Policy in all business areas that generate, create, collect, process, preserve, maintain, disseminate, or fund datasets. It is expected that making datasets available will be an ongoing (often scheduled) process and form a core part of departmental business activity.
Datasets that should be routinely published on the Data Directory include:
- datasets that are considered to be high value (see 0 below)
- datasets published in documents or reports that could be made more reusable (e.g. figures currently included in PDF reports, annual reports or reports already being provided to the Commonwealth)
- datasets which are already in machine readable format
- existing data catalogues (for example, Spatial Datamart)
- datasets currently made available on agency websites
3.2.1 High value datasets
A high value dataset is defined as a dataset that is likely to be of interest to the Victorian community, and/or a dataset that has potential for valuable reuse. A dataset should be considered ‘high value’ if it:
- is central to the department/agency functions e.g. DTF and budget data
- has been requested via ‘suggest a dataset’ (see (a)(i) 3.2.2 below)
- has previously/regularly been provided under the Freedom of Information Act 1982 8
- supports a major reporting process of government, for example, annual report data
- planning data
- spatial data
- transport data
- administrative data
- financial data
Other types of data may also be high value. Agencies should consider which of its datasets are high value. Datasets that are not high value must still be released under the Policy, but agencies should prioritise high value datasets.
3.2.2 Dataset suggestions
The Data Directory includes a function called ‘suggest a dataset’9 that allows the community to make suggestions for datasets to be made available under the Policy. Once a suggestion has been received on the data directory, it will be forwarded to the relevant agency. It is the responsibility of the agency’s dataset custodian (see Section (a)(i) 10 Accountability for datasets) to respond to the request within four weeks and if appropriate provide the relevant dataset in a timely manner.
Suggestions received on the Data Directory will be recorded in a central register on the website that will be displayed to the public. The agency custodian will be responsible for updating the register to record outcomes of suggestions relevant to their agency.
It is recommended that agencies report suggestions for datasets received, and the outcome of these suggestions, in its annual report. Agencies are also encouraged to provide a link from its agency website to the ‘suggest a dataset’ function on the data directory.
Further detail on the ‘suggest a dataset’ process including responsibilities, expectations and timeframes of responses is available on the DTF website.10
3.2.3 Establishing a process to make datasets available following freedom of information requests
Datasets made available under a freedom of information request must be considered for release under the Policy. The release of data via the Data Directory is consistent with the intent and language of the Freedom of Information Act 1982. Datasets will still need to be assessed as supporting the Policy intent and not breaching any restrictions. The time requirements stipulated under the Freedom of Information Act 1982 do not apply to the Policy.
3.3 Collections of line agency data
Departments are encouraged to support its portfolio agencies in releasing data under the policy by collating data in meaningful collections, for example, Department of Health and Human Services making data available that it collects and compiles from hospitals.
3.4 Research data
The default obligation under the Policy is for agencies to make research datasets available. The advantages of sharing research data include that it:
- encourages scientific enquiry and promotes innovation
- leads to new collaborations between data users and data creators
- maximises transparency and accountability
- reduces the cost of duplicating data collection
Allowing access to the data which underpins research significantly increases the potential benefits of the research. For example, it allows other researchers to test the findings in a research paper, or to take the research in a different direction.
Accordingly, where an agency owns or licences the data that underpins a research paper, the data should be released under the Policy if it meets the criteria for open data release.
Further, where an agency funds research which is likely to be underpinned by data, it should consider including appropriate terms in the funding or grant agreement to:
- allow the agency to release the data (see Chapter 6 Licensing datasets for more detail)
- require or recommend that the researcher provide access to the data in a form that is compliant with the policy
3.5 Creating an information asset register
Victorian Government Information Management Standards require agencies to create and maintain an information asset register. Using its registers agencies will be able to identify and manage datasets which can be made available to the public.11
3.6 Open data plan
Agencies are encouraged to develop and publish an open data plan. An open data plan describes the process for the release of an agency’s data. The strategy should include:
- a description of the agency and the types of data the agency manages on behalf of the State
- the agencies legislative and administrative framework
- a statement of compliance with the Policy
- the agencies governance and accountability framework
- a list of short and long term goals
- a plan for what data will be released and when
- a discussion on what data will not be released and why
- a list of all data already released
The supporting data to enable developing an open data strategy should be available in the agencies Information Asset Register (see section 3.5 for more information).
Publishing the agencies open data strategy should be done on the Data Directory.
11. Guidance on information management, including asset management, can be found at: www.vic.gov.au/information-management-policies-and-standards