Academically educating Scotland in data skills
The data and analytics industry is firmly established in Scotland, from The Datas Lab and its annual Data Fest through to data-driven start-ups like Skyscanner. But what of the institutions turning out the trained data professionals to work in that industry?
I spoke to the convenors of one of the long-running data science Master’s courses in the country, as well as one of the newest, to find out what it has been like to train the data professionals academically that are so in demand by the Scottish data industry and in the rest of the UK.
The University of Dundee has one of the longest-running MSc data science courses in the United Kingdom with the first intake enrolling in 2010. Andrew Cobley, senior on the course which is part of the computing department, explained that it was borne out of the Master’s course in business intelligence.
Cobley is the organiser for the initial degree course which taught skills around business intelligence, OLAP cubes as well as statistics and maths. However, it became clear to him from speaking to the students, many of whom still had their day jobs in the data industry, that change was afoot in the industry.
"The field was changing - students were asking about data lakes and machine learning."
He said: “We were talking about traditional relational databases and it was obvious the world was starting to invest in Cassandra and MongoDB and Hadoop. We could see the field was changing. The students were asking about data lakes and machine learning.”
From the students’ questions he could tell there was something in the air. It was time to create a new course that would help that to answer those questions and understand the new technologies and tools that were becoming popular in their workplaces.
The new course has at its core of business intelligence because Cobley firmly believes that in the new field of data science with its NoSQL databases, you can’t forget the analysis. He said: “People still need to know about older traditional technology because if not they might think that they have created something new, but in fact it was just something they have never heard of.”
“We knew that we would have to be agile in teaching this course."
Unsurprisingly for a course that has been developed to satisfy the curiosity of inquisitive students, the syllabus of the MSc Data Science is flexible and fluid to allow for future changes in the industry. Over the last two years, the curriculum has incorporated a lot more applied machine learning, programming languages and statistics.
Cobley said: “We knew that we would have to be agile in teaching this course. Data science is one of the quickest developing fields and it’s changed so much just in the last five or six years that we’ve been teaching it, so it is absolutely vital that we don’t stand still.”
Ninety miles away at the University of Strathclyde in Glasgow, the first cohort of students are having their final projects marked at the end of the MSc in Data Analytics. Dr Kerem Akartunali designed and convened the first year of one of the newest data science Master’s courses in Scotland. He said that for him it was important to give the students a broad base of skills.
"A lot of these courses are heavily computer science-dependant."
He said: “We have a lot of data science, analytics, big data type courses around the country and one thing that struck us was that a lot of these courses are heavily computer science-dependant and, from our perspective, they were missing some other key aspects.” The “us” Akartunali refers to comprises himself and Dominic Finn, the convenor of the course from this academic year.
He said that data science is made up of different blocks including machine learning, statistics and optimisation as well as big data for predicting and forecasting. As such, the course is being delivered by three departments, with the Department of Management Science at the Strathclyde Business School taking the lead. The other two are the Mathematics and Statistics Department and Computer Information Sciences Department, both in the Science faculty.
Finn, a teaching associate at the department of management science said that, like the University of Dundee, the lecturers benefit from the contact with industry via the students. He said: “One of the things we’ve managed to build into the course is conversations with the students.” Akartunali also added that their students are exposed to challenging practical problems so that they develop business as well as practical skills.
"We discuss with organisations how they percieve data science and analytics."
Finn also said that they took the approach of talking to organisations about how to develop a rounded data analyst that can understand the technology and the statistics and translate the business requirements. To do this, the business school, the faculty leading the data science courses at the University of Strathclyde, has a departmental advisory board that involves senior managers and analysts from various organisations.
“Over the last years, we’ve had some discussions with them to understand how they see and perceive data science and analytics and it is quite a fast evolving area,” said Akartunali.
Close contact with industry is also beneficial to the students at the University of Dundee as it gives them the opportunity to network with potential employers. Cobley said that companies often come to the university and give a talk for an hour and then go out to dinner with the students, resulting in an exchange of views and occasionally one or two job offers.
"They want to move from Excel to much bigger datasets."
However, due to the structure of the course at Dundee, with most students choosing to take the part-time programme which involves distance learning and two separate weeks of intensive on-campus study each year, many who enrol already work in the data and analytics industry.
According to Cobley, they just want to formalise what they know. “Quite often they have been working on Excel spreadsheets, but they want to move into much bigger datasets and doing much better analysis.”
He added that they also get manager enrolling on the course who want to be able to communicate better with their technical teams. For these students, there isn’t the stress of job-hunting after graduation as they will simply be promoted to a new role within the company, though some do move into more of a “data science-y job for bigger companies,” said Cobley.
Though the students of Akartunali and Finn are still having their dissertations marked, some have already found jobs, in Scotland and further afield in places like Italy and Germany. Akartunali said that six months after graduation, he will gather information on student destinations, so it is too early to give a full picture on the success of this first cohort.
However, he and Finn are pleased with how the first year of the data science post graduate course has gone, as well as how the well the three departments co-ordinated and synchronised to deliver the course. “It worked very well, the students are very, very pleased and we were also quite happy. It was quite a good experience,” said Finn.
MSc Data Science at the University of Dundee costs £7,950 for home, EU and rest of UK students, and £19,950 for international students.
MSc Data Analytics at the University of Strathclyde costs £10,300 for home, EU and rest of UK students, and £19,500 for international students.