A few months back, I was lucky enough to attend the Latin American E-Science Workshop: Turning Data into Insight. It’s run jointly by the São Paulo Research Foundation, FAPESP, and Microsoft Research, and does a fantastic job of enhancing the skills of young scientists, as well as increasing the visibility of Brazil as a centre for research.
Sounds awesome. But at first, I wasn’t even sure if I should apply. I’m working on bee populations in agriculture – it is definitely science, but calling it ‘E-Science’ sounds like a bit of a stretch. Not quite so, as it turns out.
In fact, E-Science encompasses a range of research fields and commercial sectors, with one thing in common: they use very large datasets. But remember that large is relative. My ecological dataset, consisting of maybe 100,000 rows of data, is in fact laughably small if we compare it to those of genomics. Clearly, the organisers of the workshop thought that it was big enough.
The workshop turned out to be pretty darn amazing with an exceptional group of speakers from a variety of backgrounds, including computer scientists, software engineers, ecologists and conservationists. Of course the diversity of speakers could have been cause for failure; after all, the vocabulary of a computer scientist can be completely different to that of an ecologist. But at this conference, the diversity was its triumph: every talk had enough detail to satisfy the specialist as well as key points that were applicable across disciplines. So whilst the computer scientists could engage with the intricacies of relational algebra and adaptive executions, I could focus on the simpler ideas; like the importance of keeping detailed provenance and workflows (which can be thought of as keeping track of every single step during research, from data acquisition to the production of figures).
We didn’t just sit for days on end listening to talks, however. There was also a day of tutorials on various tools developed by Microsoft Research. Drew Purves explained the use of Filzbach, a new programme that allows you to fit complex models to data; FetchClimate, which extracts climatic and other abiotic data for a set of sampling locations; and Distribution Modeller, which integrates them both. Rob Fatland introduced us to Microsoft’s data visualisation tools: WorldWideTelescope (WWT), which was developed for astrologers, and its Earth Science equivalent, Layerscape. With these tools, you can see your data in novel ways; for example, visualising earthquakes by looking up and the surface of the earth, and observing river currents moving through space and time.
The DemoFest was also an insight into the exciting software developments and projects that are currently underway. Like ChronoZoom, which is “an interactive timeline from all history”. Sliding from the birth of the Milky Way Galaxy, to the dawn of human history, really gives you a new appreciation for the magnitude of time. I admit: this isn’t directly relevant to my work, or useful in my PhD other than it’s provision of hours of fun procrastination. Nevertheless, it’s amazing to see the different ways in which people approach and solve the non-trivial issue of creating insightful and intuitive visualisations for Big Data.
After the workshop, we also had a couple of days to explore São Paulo. It really is an amazing city, and definitely worth a visit. Imagine a very shiny, bustling city full of skyscrapers. Now get some traditional, beautiful Mediterranean villas and squeeze them between the skyscrapers. Add in traffic on the ground, helicopters overhead to avoid said traffic, a 24 hour bakery and the worlds best sandwich, and you’re in São Paulo, my friend. This sandwich, oh this sandwich, was stuffed full of mortadella (an Italian meat), with a bit of provolone dolce (a yummy sweet Italian cheese), and some lettuce (the least important ingredient) in between some delicious, freshly made bread. It was bigger than my head, and I don’t have a petite head. If you do visit São Paulo, head on over to the Mercado Municipal for this scrumptious treat.
Anyway, I’ll stop rambling about food and move on to more important things… Like the many job opportunities for academics at the University of São Paulo (USP). There’s a big push at the moment to entice bright academics from around the world, a push which has been lacking in the past. It used to be that to apply for a job at USP, you had to write the application form and take an academic exam in Portuguese. So you had to learn the language before you’d even been offered a position. Tough. That’s changed now though, so hopefully they’ll see a lot more interest from internationals – it’s a beautiful country after all, with a growing research scene. And if that’s not a deal sealer, how about this: in Brazil, once you become a professor, you are essentially a civil servant with a job for life (or so I am told by an inside source). On the flip side however, this could mean less drive to continually do leading-edge research and remain prolific. Although maybe it means that they have the security to undertake more ‘risky’ projects. Something to think about I reckon.
So, concluding my long ramble with a shorter ramble, if you are interested in Big Data, E-Science, reproducible research, data visualisation, open source, data analysis, cloud computing, or pretty much anything where you use a reasonable amount of data, apply for funding to attend the next conference or workshop! Not only do you learn how people from a variety of disciplines approach and solve the same problems either in collaboration or independently, but you get to explore the research environment in São Paulo and speak to students, academics, and professionals, from a vast array of countries and disciplines.
All that is left to say is, obrigada FAPESP for giving me the opportunity to take part in this fantastic workshop!