This is my second time attending an RDA plenary. The first time was in Tokyo 2016. The first plenary for me was as much about learning what the RDA was about as it was an exciting introduction to many new data management concepts as well as an opportunity to meet wonderful new people. I feel very luck to have been selected as a grant recipient for both occasions and am very grateful for the generosity of the RDA in providing me with this opportunity to be at the cutting-edge of Research data management processes and discussions.

For the second experience in Philadelphia, I was a bit more familiar with the work of the RDA and I had become more acquainted with and had developed some opinions on concepts such as FAIR data, persistent identifiers, ethics in data science and analytics, software source code, RDA outputs and recommendations, etc. For this reason I think I found that I was able to get a bit more absorbed into the presentations and discussions this time round, as I was already introduced to matters on hand.

I think it was clear that this was the case for other attendees at the breakout sessions as well and the discussions included more considered questions that displayed a deep understanding of the subject matter and perhaps challenged the mindsets which had developed up to now while at the same time offering scope for improving things further.

I tried to get involved to some extent in all the groups I attended, but here are two brief summaries of two groups I attended over the course of the plenary. The Early Career and Engagement Interest Group (ECEIG) had a meeting on Wednesday evening. The meeting began with an introduction to what the ECEIG was about. The group arose out of a need to provide a degree of support early and mid-career professionals. The focus of this support is to provide opportunities for networking and mentoring amongst members. This is put into practice through organised social events at each plenary, a mentoring programme and ‘Ask-Me-Anything’ (AMA) conference calls.

Over the course of the meeting we heard from the organisers who talked about the aims and goals of the group and who then opened the floor to researchers to talk about their work. This was then followed by a general sharing of advice and mentoring from across the floor. This was quite useful to participate in as well as to obverve. The organisers also discussed their plans for the group over the coming months and invited active participation from attendees. The ECEIG is a useful group to share experiences with other members at different stages along the career line.

The expected outputs of the ECEIG are to:

• Establish a volunteer-based mentoring progamme

• Network across domains to establish an interdisciplinary network of peers

• Provide a space for people who have more experience with RDA to pass on knowledge and lessons learned to Early Career Researchers and Professionals

• Create a social outlet specifically for Early Career Researchers and Professionals

The meeting carried these objectives a step further by offering an opportunity for organic mentoring to occur. This was enthusiastically availed of by members old and new. There was an openness for such mentoring and this took a significant portion of the meeting. The mentoring discussion also naturally segued into a conduit for more experienced members of the RDA to pass on their experiences to newer members. It can be expected that the mentoring will continue through the volunteer mentoring programme after the plenary, as part of the AMA calls and perhaps side conference calls organised with a specific purpose, e.g. presentation skills. AMA calls may also be organised on specific topics such as machine learning and CODATA . Opportunities for networking will also arise at the conference calls, and at plenaries – in particular as a result of this environment for members will to become acquainted and this hopefully will lead to networking across disciplines in areas of mutual interest. These outputs should make membership of the RDA and attendance of the plenaries more meaningful for early and mid career researchers – firstly as they will help to break the ice and build ‘social-bridges’ between new members who might such a plenary overwhelming and isolating but in this process it should foster a better understanding of the work of the RDA. In the longer term as ECEIG members become more experienced and engaged with the RDA, they may choose to give back through either leading or becoming members of other interest groups, working groups, other RDA organisational groups such as the TAB, or simply passing on details of RDA outputs to their research environments or places of work or else actively contributing at plenaries.

Another group I attended was the Software Source Code Interest Group (SSCIG). The group aims to discuss a broad range of governance issues relating to software source code such as management, sharing, discovery, archival, provenance and metadata.

The meeting had three main objectives:

1) Share information on activities taking place in 2018 since RDA 11 (there was no meeting in RDA 12 due to the location)

This segment of the meeting included an update on the work of the Force11 software citation implementation WG. Details were also given on the launch of the Software Source Code Identification Working Group. This WG aims to produce recommendations and guidelines for software artefact identification and their adoption by relevant stakeholders, e.g. publishers, repositories, and archives. This working group is a joint RDA/FORCE 11 endeavour.

An outline was provided on the National Plan for Open Science in France. This document describes the open science policy adopted by France with the aim of making research publications and data freely available. This segment also covered efforts to raise awareness about the relevance of Software Source Code and included details on the Paris Call on Software Source Code and the Research Software Alliance (ReSA). The work of this group may be related to the Research Software Alliance. The focus of ReSA is to promote software as a vital component of research. A suggestion that arose in the meeting was the need to coordinate international efforts with national activities with ReSA. This is twinned with another identified need to raise awareness of software as a key asset. Adopteres could also be encouraged to sign the Paris Call on Software Source Code (UNESCO) which centers around the preservation of software code as a heritage, while sustaining Free and Open Source Software communities.

Copyright law was also discussed – details were provided on an upcoming EU Directive that aims to reform copyright law. The concept of what might be considered open source code is one that could affect current software development and sharing platforms. It would be preferred that the directive does not mention ‘open source’ code as it may endanger such software archiving and distribution platforms. An EU Directive must be transposed into law in EU member states, and it is possible that the details of such Directives may also be introduced by other lawmakers across the globe. Details were also provided of the software heritage archive which preserves software code for future generations.

2) Group work and discussion on FAIR for Software Source Code

There was a discussion on what it means for software to be FAIR. It was interesting to hear that the discussion on FAIR which up to now had been covered with data was moving to include software source code. The group also broke into to little subgroups who each discussed what makes software FAIR and then reported back to the group. It was useful to discuss the nuances of making versions of software as well as versions of dependencies FAIR. The following issues arose from the discussion, what was the difference between FAIR data and FAIR software, adding a ‘T’ to FAIR for trustworthy Software, how to make software findable - what unit gets an identifier? How do you deal with versioning and snapshots of code when producing identifiers ? Arising from this discussion it was suggested that the SSCIG encourages citation of software code.

3) Group discussion on the various purposes of making research Software Source Code available.

There was no time for this discussion, as there was a lot of enthusiasm in the first two items and the SSCIG ran out of time.

On the plenary as a whole, I can say that it was useful to question the outputs of algorithms – particularly as the decisions they make are now more likely our day to day lives in some way. This goes hand in hand with trustworthy data and software. It was also interesting to hear the discussion on ethics in data science which didn’t concern privacy. I would have thought that there was more scope for data scientists and statisticians to get more involved in the RDA – even if was to set up birds of a feather groups. It was good to see that the discussion was maturing and that poignant questions on what could make the RDA outputs more successfully adapted going forward.