A snapshot of open data in Canada

How many Canadian scholarly journals do you think are fully open access?

50? 100? 200?

Just last year, academic librarians Jessica Lange and Sarah Severson studied Canada’s publishing industry and identified 292 Canadian open access journals.

Yes, the open access movement in Canada is underway. Publicly funded, government granting agencies developed policies requiring grant recipients to make outputs of research openly accessible to the public. In theory, these policies are a positive change, given the view of science as a public good. In practice, for authors to comply with these policies there can be cost-related barriers encountered.

Open access is one unit of open science. It isn’t only the published reports of research that are undergoing an open-revolution. The data supporting the research are too.

In Canada, open data efforts span multiple sectors. In academia, researchers are both using open data and creating open data, prompting the development of data papers that describe datasets to help ensure they are FAIR—Findable, Accessible, Interoperable*, and Reusable. The NGO DataStream Initiative created platforms to share information about freshwater health about the Great Lakes and other Canadian watersheds. Multiple levels of government in Canada are also embracing open data. The country’s latest National Action Plan on Open Government sets out milestones to build an “open data ecosystem” as part of its larger Open Government Strategy.  

FAIR Data Principles | Open Science Training Handbook (CC0 1.0 Universal)

Canada’s commitment to open data

Providing over 30,000 datasets through the Open Data Portal, federal government-level datasets cover myriad topics, such as, the mother tongue of residents, the Ontario Tree Seed Transfer Policy, and Canada’s War Dead Honour Roll.

“The Government of Canada is a leader in open data among Organisation for Economic Co-operation and Development countries, generating and releasing troves of valuable research data,” write open science specialists Dominique Roche and Monica Granados in their collaborative perspective in FACETS on open government data in Canada.

Wilfreda Edward-Dolcy, Open Data Specialist with Canada’s Open Government Team, sees open data as a tool “that enables us to build citizen trust, strengthen our democracy, harness technology by bringing new voices into policy design and public service delivery, and foster better community outcomes.”

Work by journalists demonstrates some of this functionality. The Toronto Star’s Open Data Team publishes daily updates on where road closures will be in and around Toronto. These articles are “automatically generated using open data from Ontario 511,” an information service of the Ontario Government’s Ministry of Transportation. Meanwhile, CBC/Radio-Canada journalists used open data to raise awareness of the underrepresentation of women candidates during Canadian elections and highlight trends in winter warming in every major Canadian city.

From secure by default to open by default

While increasingly being adopted across Canada and elsewhere, open data is still a relatively new affair. Edward-Dolcy describes the journey as one from “secure by default” to “open by default.”

“Secure by default assumes all datasets are first and foremost protected and not disclosed, except for select things that we identify as appropriate for publishing out in the open,” Edward-Dolcy explains. “Open by default assumes everything should be released publicly while ensuring privacy and security requirements are upheld.”

For example, survey responses on climate anxiety in children and young people are available as open data because the dataset is anonymized to maintain privacy and hosted on a reputable data repository to maintain security.

Indeed, open by default doesn’t mean privacy and security, alongside legal and ethical requirements, should be ignored in the quest to create open data. Rather, open by default recognises what Edward-Dolcy references as the “spectrum of openness” and the Open Data Institute as the Data Spectrum: along a continuum, data can be made available to individuals, certain groups, or anyone.

The spectrum of data openness | Open Data Institute (CC-BY license)

An interesting use case of this continuum is the release of data on the locations of at-risk or sensitive species. Unobstructed access to this data could lead to harmful misuse (e.g., poaching). Keeping the data “closed” and unavailable to responsible stakeholders could hinder biodiversity research or effective conservation action. The Global Biodiversity Information Facility (GBIF) recommends transforming the raw location data into less precise coordinates as one way to facilitate sensitive data being shared openly.

Respecting community rights in an open data landscape

It isn’t just sufficient to make data open. To realize the societal benefits of open data, the data must be representative of the diverse peoples and communities it is  meant to benefit.

“Representation in open data is essential because it allows us to look beyond the data to find the stories in need of being told,” says Edward-Dolcy. Crucially, “this means that data subjects must be consulted to contextualize the data that is about them. Through this community buy-in, we raise open data awareness and ensure that communities see themselves in the open data-driven stories we tell.”

Consulting communities and considering their needs is not yet a standard practice of Western science.

Community-engaged research (CER) strives to advance community goals through collaboration between researchers and communities.

Eve Tuck, Indigenous scholar and Associate Professor at the University of Toronto, views collaborative approaches as “a way to repair the legacy of harmful research practices.” Such practices include extractive data collection or reuse of data without consent. Neither practice upholds the rights of Indigenous Peoples to govern the collection and application of data about their communities (i.e., Indigenous data sovereignty).

In Canada, the First Nations Information Governance Centre created the OCAP principles of Ownership, Control, Access and Possession.

These principles are essential to follow when working with Indigenous communities, especially in an “era of big data and open data.” As Edward-Dolcy emphasized, open by default is not a careless movement, it is one mindful of ensuring human rights.

By all stakeholders of data working more closely and considerately together, the future of data won’t just be more open. It will hopefully be more just.

—

*Interoperability refers to having metadata in a common format, with terminology and vocabularies following industry standards.