The Big Data and also Machine Learning conference in Rome has finished on last Saturday, so I thought I’d write some observations, thoughts and perhaps troll a bit (a troll a day keep the doctor away).
First a bit of a disclaimer – it’s my first non-purely developer conference, so some of the things that were odd to me, might be totally normal in this space. Moreover, maybe me not being a native Italian didn’t allow me to appreciate all the presentations and conversations to their full potential.

IMG_20170225_145916

 

Organization

So let’s start at the basics – the organization. I think that folks from the Codemotion crowd show that the experience is the best teacher and they pulled it off really nicely – almost no glitches if not for some occasional sound/microphone issues. But for that I guess you’ll have to blame Murphy. Ok I might point out that single panel “Big Data: Big Capabilities, Big Concerns”, forcing all the attendees in the hall that was clearly too small – could be viewed as a bit of bummer. But the solution would be to find a bigger hall/different location (and Roma Tre has proven to be a very good space also before for Codemotion), or limit the ticket “sales”. So I can totally forgive a minor inconvenience, for the cool conference with killer price.
Another thing that also falls under organization is the “batch” approach to sessions, which I haven’t seen before. Usually there’d be a synchronized talks (starting and finishing at the same times) in all the rooms, with about 5min break if you’d want to switch. This choice of machinegun-presentations surely allowed to cram more content in the time slots, but it wasn’t as easy to cherry-pick presentations. Tradoffs, tradoffs
 
IMG_20170224_164246
 

Presentations

I won’t be able to talk about every presentation I attended, still I have to point out and praise a few presentations that were informative and in my opinion worthwhile. First theMachine Learning Real Life Applications By Examples” should have been titled “Introduction to Machine Learning: History, and why should I care”. The content was unexpected judging by the title, but still it was a well executed presentation. Second “Il deep learning ed una nuova generazione di AI“ – where we saw state of the art AI solutions, and the new risks that they bring with them (like racial bias, or our expectation that AI should be smart while it’s still generally stupid). And lastly the “Clustering Reddit: Learning from The Front Page of the Internet” where as advertised – Sara di Bartolomeo showed what happens when you throw a Machine Learning book, with all it’s algorithms at a ton of data which is Reddit (aka frontpage of the Internet). Worth the mention is a cool idea of extracting sentiment analysis material from the Amazon’s comments.
 

General notes

The crowd was very diverse – much more so than the Ruby on Rails conference that I’ve attended the last few years. Also even though the subject of the conference wasn’t simple, there were lots of young folks in the audience (and I’m talking collage student level young). Good, good. Another thing worth pointing out is the fact that it’s the first time where I saw this many, and this openly advertised job positions. While sponsors most often fund the conference to gain advertising space, and notoriety/prestige – it’s not uncommon to use it as recruiting opportunity. Still never did I saw it made on such a level – just an interesting data point.
But in the end I did enjoy the conference as a whole. It was a great opportunity to meet people with similar interests. Also, a bit more of a personal reaction to this (and other conferences) is “DAMN! there are people making all these cool stuff – time to do some too!”. Even if just for that it was definitely worth going to it. Oh I would have forgot – FREE FOOD AND BEER ON COLLEGE CAMPUS.
So I’d like to thanks to organizers for all the hard work – they did an amazing job. Thanks to the speakers for all their effort, and the sponsors that made it possible for the event to happen. I hope to be there again on Data Driven Innovation 2018.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes:

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>