Data science is undoubtedly one of the hottest careers around at the moment, but there are ongoing concerns about a shortage of skills in the field. A recent report from Crowd Flower revealed the size of the challenge, revealing that a staggering 83% of respondents were struggling to find people to fill vacancies in data science related roles.

Of course, this is not a new trend, with Gartner highlighting the issue way back in 2012, but the Crowd Flower data suggests things are getting worse, not better.

I’ve written previously about attempts to make data science easier to perform by using AI to automate much of the technical side of things.

Supporting collaboration

Another way to get more out of the data scientists that do exist is to encourage and support greater collaboration between them.  That’s the mission of Data.world, a new Texas based venture that has recently launched a social network for data scientists.

The platform comes with around 1,000 datasets already uploaded, and hopes to fundamentally change how research is performed.  They work hard to ensure that all of the datasets are machine readable and easy to dive straight into.

Before Data.world, it was much harder for data scientists to get together and work collaboratively on the same dataset.

The process is designed to mimic more mainstream social networks, with users tasked with filling out a profile, complete with any data they want the community to work on.  The algorithm underpinning the site will then match them up with potential projects they’d like to get involved with.

At the moment the site is in a beta phase so is accessible by invitation only, although a public launch is scheduled in for later this year.  The site is free to access, with all public data included in that membership.  If members wish to gain access to private datasets then a fee is applicable.

The team behind the site are confident that getting data scientists together like this can fundamentally change what the profession is capable of achieving.

Of course, there have been crowdsourcing sites, such as Kaggle, around for a little while, so time will tell just how effective Data.world is at differentiating itself.