Towards NeIC2019: Interview with Christine Kirkpatrick

NeIC2019 takes place at the Tivoli Hotel & Congress Center in Copenhagen, 14-16th May. One of the four Keynote spekers is Christine Kirkpatrick, who represents the San Diego Supercomputing Center. At the conference, she will be talking about Open Science in the US. We interviewed her to get a head start on the topic.

Why are you attending NeIC2019?

I am honored to be an invited keynote. As well, I enjoy the Nordic infrastructure and data community. Since I first connected with the group in 2016, it’s been a source for pragmatic, forward thinking research services. I attend for the intellectual vitality and the welcoming, kind approach my Nordic colleagues have shown.

How do you benefit from Open Science and data sharing in your own work?

I have not benefited as much as I’d hoped. I spend a lot of time, as do many researchers, in trying fruitlessly to gain access to existing data (if I’m lucky enough to find it), gathering, and ultimately creating my own datasets. I could get further into inquiry, faster, if I could skip this step and find my way to FAIR data in my area of study. For example, my research at the moment is centered around optimizing cloud architectures for data science use cases. Many researchers just use the same attributes/resources as when the analysis ran successfully before, rather than tuning to the needs of their applications, or to converting into pipelines that make use of heterogenous resources, such as HPC and cloud. It’s turned out to be much easier, albeit time consuming, to study local resources where I have access to logs. We’ve run into similar issues with social media analytics research. We spend an inordinate amount of time screenscraping data, and ingesting data such as Twitter via APIs, when the data stores exist – in a FAIR, cleaned up state, but that we can’t get to because of economic, proprietary, and/or licensing issues.

What would you say is the most important reason to promote open research data and science?

I am motivated by the concept that we have much of the data we need to solve society’s great challenges, we just can’t locate the right sources, and haven’t figured out how to put it together. The complexity and vast availability of data means we need more help than ever from machines to mine data for insight.

In your presentation, you’re talking about implementing the models used in the EU to the American field of Open Science. Are there a lot of differences to consider?

The biggest differences are sociotechnical, which results in big technical approaches. The EU is well funded by comparison, shows unity in its commitment to open data, and has established policy to ensure progress is made. The US by contrast has modest results from a patchwork of funding agency and publisher policies, but completely lacks a unifying framework like Horizon 2020 or a project that transcends agency boundaries, such as EOSC.

What would you say are the main challenges in Open Science in the US?

At present, I find our labeling of ‘open science’ and overuse of ‘FAIR’ as solutions and destinations to be an impediment. Don’t get me wrong, I am grateful they are well represented in scientific discourse. But these words only speak to those already in the tribe of open science promoters, that is, those already convinced. We need to find new ways to approach researchers, industry partners, funders, publishers, and support staff to be motivated to contribute and realign practices. The first step is framing ‘open science’ instead as an issue or challenge that the stakeholders recognize and own. For example, we know from several sources that most of a data scientist’s or data analyst’s time is spent on finding and wrangling data – tasks that neither stimulate these workers, nor make the best use of their finite time. If we spent time recasting open science as a way for industry and research initiatives to get deeper into a subject, more quickly, with more results – or as a way to save money on data analysts, this might bring immediacy to audiences. This will draw new participants into the conversation and build momentum needed for greater adoption and impact of open science techniques.

What is your view on the EOSC-Nordic project?

As a member of the external advisory board, I am extremely excited and optimistic about the impact the community will see from the EOSC-Nordic project. Gudmund Høst assembled a stellar team and a smart approach for coordinating across the Nordic and Baltic regions.

Please find more information and register for NeIC2019 here. You can still register today and tomorrow. We hope to see you in Copenhagen in May!