​​​​​​​​​In the rapidly evolving digital era, with a plethora of data being generated every second, the management and access to data have become increasingly challenging. The 18th International Open Repositories Conference (OR2023) hosted by Stellenbosch University’s (SU) Library and Information Service, yielded fascinating insights and possible solutions to some of the most pressing information management problems of our time.

The OR2023 conference was held from 12 to 15 June at the Lord Charles Hotel in Somerset West and attracted close to 300 experts from 41 countries. The conference featured speakers from a wide variety of sectors who engaged about cutting-edge technologies, innovations and best practices in data repositories. It was the first time the event was hosted in Africa.

At the official opening of the conference, Prof Sibusiso Moyo, Deputy Vice-Chancellor: Research, Innovation and Postgraduate Studies at SU, delivered an opening address. She described OR2023 as an excellent platform for library practitioners, scholars and researchers to exchange ideas on how to effectively preserve and share research data for better discovery and interoperability. 

“In the higher education sector and as research-intensive universities, we know that libraries have had to evolve with the times, not only as places where we go to find books to read but as strategic partners, providing innovative solutions for our research, innovation, teaching and learning missions,” Moyo said. 
The theme of OR2023 was “Repositories unlocked for discovery and interoperability” and included a number of sub-themes of which two are particularly significant in the African context: “Unbiased discovery of knowledge: Open repositories facilitating the unbiased discovery of indigenous knowledge” and “Sustainable development: Linking open repositories content to the UNESCO Sustainable Development Goals (SDGs)”.

​The opening keynote panel on the second day was facilitated by Claire Knowles from the University of Leeds in the United Kingdom who challenged participants to think about how repositories can improve data equity worldwide and how they can be better utilised to share data with both humans and machines. 
SU’s Senior Director of SU Library and Information Service Ellen Tise explained that the rationale for open data is similar to the goals of other movements such as open source, open education and open government. “Open data is accessible, exportable and could be shared by anyone for any purpose. If you look at data through the equity lens, we should consider the ways in which data is collected, analysed and interpreted. It raises concerns regarding power and privilege, knowledge and data sovereignty. If we don’t have access to reliable data, we can’t make good decisions.” 

Tise argued for deliberate steps to remove barriers to data access and facilitate participation for marginalised communities. She said SU is a pioneering institution in the field of open access in Africa and SU’s open repositories are accessed from every continent, excluding only Antarctica. 

Equity is a key issue to consider in the context of access to information, said Dr Tshiamo Motshegwa, Director of the Open Science Platform. He emphasised the importance of data literacy so that people are empowered to engage with data in the unique ecosystem repositories provide. “We need to ensure data reflect diversity without bias and exclusion. We must collaborate on the African continent to utilise open science to solve societal challenges, but we should also strive to collaborate globally. Perhaps we need a World Data Organisation, similar to the World Intellectual Property Organisation, to facilitate data policy, services and cooperation.”

Joy Owango, founding director of the Training Cente in Communications in Kenya, reminded the audience that equitable access to data should not just be focused on the research community and academics, but also serve the broader multi-sectoral community. “Our challenge is strategic awareness. Let’s look beyond our community and ask what other sectors are looking for and how we can make ourselves more accessible by taking advantage of technological developments.” 
Owango urged repository and data managers to become more involved in discussions about open science and policies. “We’re not seeing those responsible for managing and sharing the knowledge in strategic conversations about policies. You should make yourselves more accessible so you can share your concerns with the powers that be when it comes to data regulations.”

Open repositories are a resource for the global public good, argued Prof Willem Fourie, founder of the South Africa SDG Hub. He explained that the SDG Hub is a meta-platform aimed at making research available to sustainable development policymakers. “We work with metadata from 14 local repositories as well as highly-ranked universities from abroad. More than 160 000 research items from the past ten years are classified through machine learning in terms of one or more of the SDGs.” 

Machine learning can help make repositories more accessible and can add value in many fields, Fourie said. He added it’s fair to assume that open repositories were part of the training data sets used in the creation of large language models (LLMs) and that research data contributed to more intelligent, accurate and ethical artificial intelligence used in chatbots such as ChatGPT. Fourie explained a new digital tool that would enable a policymaker to upload a policy and based on all the available data in a large database, identify potential areas of concern in the policy. The policymaker would then be provided with links to the relevant research offering solutions for specific issues. 

After the daily academic discussions and workshops, OR2023 delegates were treated to a welcoming reception on Tuesday 13 June in the Endler Hall. Prof Wim de Villiers welcomed delegates during this event and the South African Youth Choir performed.  The gala dinner of the Conference was held at Neethlingshof Wine Farm on Wednesday 14 June.