I became fascinated with big data after having been introduced to a few wonderful applications. One of them is known as Watson, a computer developed by IBM that can both read and understand natural language such as scientific articles. Watson is extremely powerful: a real big data cruncher and former world chess champion.
The machine also managed to obtain its degree in medicine within several months. A number of US hospitals use Watson to advise doctors on suitable treatment methods for cancer patients. In one case, they asked Watson for advice on an Asian patient. Although the doctors unanimously recommended treatment method X, Watson opted for Y. Why? Watson had read a magazine article explaining that Asian patients carry a specific gene that immunises them against treatment method X. With some 100,000 newly published scientific articles, it's no wonder that doctors don't have time to keep up: those that recommended treatment X can hardly be faulted for their shortcoming. Of course, Watson is not used to replace actual doctors. The machine suggests a number of possible treatment methods, and estimates their chances of success. The doctor then makes the final decision in consultation with the patient.
Another good case in point can be found closer to home. Scyfer is a company developed by our own Venture Lab. The company analyses photographs using big data technology. These technologies have now evolved to a stage where they can verify that an MRI scan does not contain any abnormalities. This is obviously a wonderful development, as it enables radiologists to focus on assessing photos that do contain abnormalities, saving a great deal of time and money in the process. Scyfer is also applying this technology in the steel industry. A major steel manufacturer takes a photo of each finished product at the end of the manufacturing process, and uses the software to analyse its compliance with quality requirements.
Still, I won't push my daughter to study data science unless she really wants to. If she does decide it's for her, I can guarantee she will find the work fascinating. What's more, she can certainly expect to make a good living: data scientists are extremely scarce. The National Think Tank estimates the Dutch shortage at some 8,000 people by 2018. This scarcity is affecting universities and industry alike. Universities are having great difficulty interesting graduating students in positions as researchers or lecturers. The problem also presents a challenge for the business community, with the possible exception of renowned e-business brands such as Google, Apple, Amazon, Facebook and LinkedIn, which are all extremely appealing to job-seekers. Traditional businesses are having much more difficulty establishing big data divisions and attracting the necessary expertise. This development is affecting these businesses at every level, from executives – administrators with the ability to formulate and manage a data science strategy – to hands-on data scientists – experts with the ability to carry out the data analyses and build applications – to boundary spanners capable of translating business opportunities into concrete data science projects.
So why is this the case? I believe there are several reasons. First, data science is not taught at secondary schools. This means pupils considering a degree programme don't really have any conception of the subject. Second, the number of academic degree programmes in data science is still limited. Furthermore, the existing programmes are far too monodisciplinary for my liking. Information scientists learn about coding and artificial intelligence. Econometrists learn how to work with structured data, but never get to grips with natural language or images. Business professionals learn about the managerial applications, but never learn about the underlying technologies. So what can we do to address this problem? The answer lies in three key measures. First: Add big science to the secondary school curriculum. Second: develop multidisciplinary degree programmes along the lines of our Big Data MBA ( www.mbabigdata.nl) and Analytics Academy. Third: organise more intensive collaboration between universities and businesses in the areas of both research and education. Among other initiatives in this area, we have established the Big Data Alliance ( www.bigdata-alliance.org). Businesses and universities will have to work together if the Netherlands is to play a truly significant role in the field of data science and business analytics.
I should point out that my daughter is only four years old. Although it will be a while before she starts studying and coding, I've already bought a few nice games to help her learn how to program. One features a dinosaur that answers her questions by means of a cloud-based connection to IBM Watson.
Dean, Amsterdam Business School
Big Data MBA & Business Analytics Programme Director