Strata + Hadoop World wound up recently in San Jose and it was all about data. Big, real time lakes of the stuff. None of us got to go but Victoria’s Twitter feed basically got swamped by people saying how great it all was using the hashtag #stratahadoop. So I thought I’d have a look through the #stratahadoop Twitter stream and see if I could curate things down to a reasonable blog. So, here goes.
The best place to start with this is always the keynotes. O’Reilly.com is very helpful in putting them all in one place for us here.
I have been interested in the idea of Artificial Intelligence since I was a child and people have been talking about it at least that long. So, I had to check out Adam Cheyer talking about Building practical AI systems. My takeaways from this were:
Building a machine that can think like a human:
- is hard and what that means is changing as our expectations change
- is only a part of the problem the other part is the world has to be ready for what you produce or it’s a bit useless
- is dependent on you having an application, this might be something as simple as finding the perfect boot
Cheyer’s idea Viv might just have us all taking to our phones as assistants rather than talking to each other on our phones.
At university a long time ago I studied conflict so I had to check out Machine learning for human rights advocacy: Big benefits, serious consequences because states that the community of data scientists can make a serious contribution to the process of establishing legal responsibility for war crimes. My big take ways were:
- Cold hard numbers can help us tell what are very emotive stories.
- For a data scientist data is data, this is a good thing if you want an actionable result.
- A good data scientist can estimate what they don’t know. It’s an estimate but it is important.
- Presentation to a war-crimes tribunal requires a high degree of rigour.
- There are are amazing people out there changing the world with data science.
Of course, there are plenty more these are just two of the ones that I thought covered interesting topics in an engaging way.
For James Kobielus, Big Data Evangelist at IBM, the big take way theme was the rapid maturation of the open analytics industry. Being from IBM James sees this evidenced in announcements by IBM about which will make it easier to work with Hadoop and Apache Spark along with the Open Analytics Ecosystem … inagurated by IBM. He also got the piss taken out of him by a comedian, you can scroll down now to see the ‘keynote’ of that.
Nik Rouda says that for him the big deal is that Spark is gaining mindshare. That means people are talking about it, not that they are buying it. To be honest the rest of his post reads like what I’ve been hearing for the last two years. Cloud and machine learning are getting hotter, the ‘traditional’ data warehouse is over and people like information presented as pictures, not machine code. Maybe this is the year it really changes but I heard that last year.
Bill Davis over at SAS sat down with Keith Renison, a senior solutions architect at SAS. They are very excited about SAS Data Loader for Hadoop and SAS Visual Analytics because they are able to visualise large amounts of data. Renison says he is often told that 80% of analytic work is preparing data for visualisation. The analysts I work with say most of their work is preparing data, for analysis but I guess it all depends on your perspective.
For Murthy Mathiprakasam the biggest deal was the fact that customers are beginning to realise that Big Data won’t reach its potential until security and governance are addressed. While this isn’t a going to set off fireworks for anybody it is good to see someone with their feet on the ground.
Finally, Nonsense science by Paula Poundstone is a comedian talking about how much she hates technology. This has nothing to do with data or technology. But it is interesting to hear from someone at a technology conference who seems to have no interest in it.
- If you invite a comedian to your conference expect them to make fun of you.
Of course, there is a lot more content around #stratahadoop than that but I hope that has helped you wade through the lake of data (see what I did there) and see what was going on.
What was your biggest deal from the conference?
Success is opportunity meets preparation – Jack