Data Scientists Steal Headlines but Eclectic Skills Needed: Big Data twitter chat recap

big data skills
Big Data is big talk in business but few are thinking long term about how to source the correct mix of skills to make it a sustainable strategy. Tweeters at the #ITValue chat remarked that Big Data requires an esoteric mix of skills ranging from specialist coding, statistics to machine learning. Such a broad range of skills cannot be sourced in one person and a collaborative effort and leadership are a must. More than anything, the mind-set of the data scientist must embrace the scientist and the artist.

Required skills

The need for a range of skills to take advantage of big data, especially in the long term, was pointed out by @KirkDBorne who said:

@HelenRBeckett agreed and added ‘analytical modelling techniques and some specialist coding such as R’ to the list of desirable skills.

A requirement for such a range of skills would need multiple personnel to fulfil as @RobertsPaige said ‘Hard to find all those skills in one person. More likely to find overlapping skill sets, forming a team’. She went further and highlighted the importance of a big data project being a companywide endeavour:

This was agreed by @katweasle who said ‘we will need collaborative multi-disciplinary teams’. Although agreeing @HelenRBeckett pointed out that this cooperation can be hard to come by sometimes with the reasons being, not least of which:

@KirkDBorne  agreed saying, ‘Terminology is scary & troublesome for many! We need to do better’.

Regardless of the terminology, as @Clearswift rightly points out, there needs to be a combined strategy:

 

Acquisition of the skills

One of the most pressing questions which faced the tweeters was where these skills are going to come from?  It was generally felt that while many efforts were being made to produce a workforce with the necessary mind-sets to be most effective in a big data environment there was still much to be done.

The need to ensure the ‘blending of art and science in [education]’ put forward by @RobertsPaige was roundly agreed and lamented as something which is not happening enough today.  There is a need for ‘focus on deep long-lasting skills, habits of mind, & literacies’ to give graduates a longevity in the data science industry beyond whichever application or language held sway at the time, as was pointed out by @KirkDBorne.  The current prevailing education model, which @katweasle referred to as ‘an industrial education model’, did not provide an appropriate environment for these wider skills and while on-line learning programmes and MOOCs were offered as alternatives it wasn’t felt that the correct balance had been achieved yet.

 

Making use of the skills

In order to make the acquisition of all these new skilled workers worthwhile of course there will have to be business sign off for a big data project. In addition to aligning strategies and ensuring that the various areas of a business are aware of and capable of playing their part, the project has to have the financial backing of the board.

@christianve saw this as a significant challenge saying, ‘The most difficult I’ve seen is actual use cases, people understanding the value they get’.  By way of a suggestion @RobertsPaige said that it is ‘best to start [with a] goal’. Once a goal has been established it gives all participants something to work toward and the benefits which will be gained stay within clear sight.

The basics to concentrate on when pitching a big data project to the board were given by @eddieshortalton:

This was emphatically agreed by @KirkDBorne.

This is, of course, only a summary of the chat and many more issues were covered during its course. To see the full chat search for #ITvalue on Twitter. Thanks to all participants.

There is a #ITvalue chat every Wednesday at 4pm (GMT) and please do join us next time to share your views and experience.

More Twitter Chat Summaries on a variety of IT/Business issues