Reflections from BBSRC/AHRC Visualisation Conference

I spent a couple of days last week in Bristol at the Biotechnology and Biological Sciences Research Council (BBSRC)/Arts and Humanities Research Council (AHRC) Workshop on “The Challenges of Visualising Biological Data“. It was a great privilege to have had my expression of interest converted into an invitation to this (possibly) one-off event and I wanted to capture some headline reflections.

Background

The BBSRC and AHRC are two of the seven major bodies that make up Research Councils UK, funded by the Government but existing as Non-Departmental Public Bodies. Research Council funding is only awarded to the very highest quality research and so, by extension, they are very influential and powerful bodies. For them to recognise the growing emergence and importance to research of visualisation is a very positive and welcome development, signalling a commitment to promote and exploit contemporary visualisation methods for analysing and communicating research activity.

The objectives of the event were outlined as follows:

  • To showcase current activities, tools and technologies relevant to the visualisation of biological datasets
  • To encourage cross-disciplinary collaborations aimed at encouraging the development of creative visualisation tools for the biosciences
  • To explore the opportunities and bottlenecks associated with visualising biological datasets
  • To provide an opportunity for researchers across the biological, computational and creative disciplines to explore novel ways of obtaining maximum benefit from their research, whether this is by exploiting their biological data or by providing new avenues for the use of IT and creative tools

The intended output of this event will be a public available report articulating the thoughts, discussions, ideas and recommendations to emerge from the two-day event. However, the collaborative connections and relationships that may emerge from the event will also be a definite positive outcome.

Key Observations

The Buzz – I have spoken recently about my promotion of the Strata conference and one of the key aspects of this backing was my liking of the potential make up of delegates coming from a spectrum of disciplines and organisational domains. The same applied to this event with scientists, creative researchers, computer scientists, academics from the humanities fields and information scientists brought together under one roof.

A varied delegate mix is absolutely essential to ensure a conference has legs (during and after), encouraging people to step out of their domain expertise, meet with others, socialise, exchange thoughts and ideas. This unique type of energy (a curious journey through mutual suspicion, territory-crossing caution, timid connection, cross-fertilisation and finally inspiration) is almost impossible to achieve to the same extent when you only have a single group represented, which can happen with many conference occasions. The poster and demo sessions, in particular, were very loud, busy and lively affairs with energetic discussions taking place throughout the event.

Presenters - It is impossible to expect a gathering of presenters and subjects to all make a connection with all delegates and so it is not intended as any kind of criticism when I say I only really derived value from about 50% of the presented material. Typically, this was when the presentations touched on issues of theory around visualisation.

In particular, I was really delighted with the talk given by Prof. Sara Fabrikant, Head of Geographic Information Visualization & Analysis Unit at the University of Zurich, which contained many research themes relating to how users interact with visualisation on both aesthetic and performance levels. Special mention also to the opening presentation from Prof. Katy Börner, Professor of Information Science at Indiana University (amongst several other roles).

I was slightly surprised not to see the Arts and Humanities side of the event represented by a speaker. I understand it was largely a BBSRC led event but it would have been refreshing to see a non-science focused presentation at some stage.

I felt there were several occasions when an example of visualisation was exhibited in a positive light, when in actual fact they demonstrated several characteristics of flawed design. One in particular stuck in my mind: a tree diagram with the final nodes represented by pie charts of varying radius which failed on quite a few levels. On the plus side, I was impressed that there was consistent recognition of the different roles visualisation plays in exploring/analysing data vs. communicating information.

The Group Sessions – For me the highlight was the achievement of the two group discussion sessions where about 8 random (or not so random!) people were thrown together and asked to work as a collective discussing a handful of visualisation-related problems, issues and responses to statements. Rather than being paralysed by the clear perceptual differences that existed, these sessions seemed to flourish with the ingredients each person brought creating a great recipe to feedback to the rest of the conference.

The first session explored issues around complex data, how visualisation can aid the analysis and communication of it, and the second focused more on what makes a useful visualisation tool, what would be its key features etc.

The only shame is that we didn’t have chance to spend longer focusing on the key skills required in visualisation because I feel these are possibly more pressing than identifying specific characteristics of desired visualisation tools. We need to equip our scientists and researchers (both established and upcoming) with the visual thinking capabilities and broad design knowledge that helps them understand the key principles of visualisation, how the brain most efficiently processes information, why a multiple series of pie charts is not an effective method of displaying data etc. As Stephen Few remarks, these skills are not intuitive and must be learned.

There exists a pyramid of capability requirements: the ‘masses’ requiring visualisation basics, a smaller population require these taking to a more advanced level and then you move towards a smaller, upper group seeking the capabilities to author or knowledgeably commission their own visualisation tools and solutions.

I would love to see, as one of the recommendations from this session, a focus on developing a learning & teaching provision across all levels and domains of academia that equips people with the methods and inspiration to pursue better visualisation practice. Hopefully, this increased knowledge would permeate beyond visualisation alone and encourage an improvement in the design of presentation slides, which I have to say seemed to tend towards the rather bleak bullet-pointed deck at this event.

Overall, a very good event, well run by the BBSRC/AHRC, I was delighted to have been there and I keenly await the final report and the potential of future opportunities in this field.