What is a Data Scientist?

I went to an excellent seminar called ‘Advanced Analytics‘ organised by the OR society.  Really excellent day (aside: Advanced Analytics = predictive analysis rather than historic data analysis). Lots I could talk about but I wanted to pick up something that Jonathan Baston from IBM talked about. He talked about the new term ‘Data Scientists’ for those who are analysing data. So what are Data Scientists?

I tried to write down Jonathan’s definition but I only got the first two, but think I remember the last two. They are:

  • Technical Expertise, i.e. deep knowledge of some sphere of data manipulation or mining.
  • Curiosity, i.e. keeps asking ‘why’ and finds puzzles a challenge.
  • ?? mathematical/statistical skills
  • ?? Team working

A quick look round the web gives you plenty of opinions. Here are just some of them.

  • This from the Guardian Newspaper, which is itself a great data evangelist (click on the picture for a great infographic)
  • This good definition here from Michael Rappa, Institute for Advanced Analytics in the USA
  • This useful piece that says “data analyst + research scientist = data scientist”

Interestingly another speaker at the Advanced Analytics seminar, Tom Hebbert of Tesco,  sheds some more light on the subject in his talk. He described his team and he said they generally had a mathematical background, but often their previous job was in IT/software for start-up companies. He specifically said that people doing IT for large corporates was not a good fit.

So its seems a Data Scientist needs a mixture:

  • In-depth technical and mathematical skills
  • A strong business focus
  • Flexibility of mind and spirit
  • Likes working with others to solve a problem

Are you a latent Data Scientist? Well maybe you would like to look at Kaggle, which runs competitions relating to predictive modelling with cash prizes for the solution that achieves the best result. Have a look.

Jon Smith
Selective Analytics


Leave A Reply