I came across the University of London’s MSc Data Science program towards the end of 2020. At the time my Dad was fighting off Covid – and because I had also been exposed, I was quarantined with him and my Mom for three weeks while we waited to see whether we might also have contracted it. It was a strange time: a lot of anxiety over whether my Dad would be okay, lots of barrier nursing protocols to learn while looking after him, but also a bit of boredom in not being able to leave the house or interact with anyone – which is how I found myself browsing academic websites one evening. I had been searching for ways to further (and formalise) my data science career for some time, but because I come from an unusual academic background (my undergraduate degree was a Bachelor of Music), and because the program I chose would need to allow me to continue working full-time, I had been unsuccessful up until that point.

The University of London program was different: studies are 100% online and you can study from any country in the world. You also have 5 years in which to complete the degree which makes balancing study, work, and family life a lot more viable: you can go as fast or as slow as you need to. In addition, if your undergraduate degree was not directly relevant to data science you are invited to complete a Coursera course as an entry requirement to assess your readiness to participate. This sounded perfect to me: all I needed was a chance to prove that my self-taught skills were up to this challenge.

I remember coming home just after Christmas and telling my husband (rather sheepishly) about my latest mad-cap plan. He didn’t even try to talk me out of it (!) in fact he was super-supportive. So I applied, and was overjoyed, and also slightly horrified, when I was accepted a few weeks later.

I decided to register for just one semester and “see how it went” – I kept telling myself “You can stop anytime if this doesn’t work out” which I think is how I tricked myself into actually starting at all…

How is the program run?

The degree is completed through a combination of module subjects and a final project and thesis. Compulsory modules included Statistics, Python programming, Machine learning, Big data, Data visualization and Data science research topics. A variety of elective modules are available, depending which direction you see yourself going in – I chose Neural networks, Natural language processing, Artificial intelligence, and Social networks and graph analysis. Because I am very interested in both NLP and graph theory I chose a project at the intersection of these two topics for my final thesis: Automatic knowledge graph construction from news.

Each subject has a module leader who provides overall academic direction. Content is a mixture of video lectures, reading material, ad-hoc assignments and formal assignments. Students are supported by online tutors who provide answers through forums to queries about the material, assignments, and any other academic issues that come up.

Module evaluation is through coursework and / or online examination, depending on the subject.

What was it like?

If you are considering doing this programming you would probably want to ask me at this point “What was it really like?”. I would have to answer “The program exceeded my expectations, but not without some caveats.”

Lecturers are from diverse backgrounds – some are gifted researchers working in fascinating areas, but are not necessarily gifted communicators. Others are gifted teachers and have that knack of igniting interest and understanding effortlessly. In either case there was tremendous value to be had from exposure to their subjects, ideas and approaches – but at times it did feel like harder work than others! In most cases I found the module leaders to be approachable, interested, and helpful.

The lectures themselves do not necessarily go into laborious detail on every single concept! Lecture content and recommended readings should be viewed as a syllabus providing direction – additional reading and research is most definitely required and expected to deepen knowledge. Some students were a bit miffed by this, but at masters-level I was expecting to do quite a bit of independent digging and research so I was fine with it.

Studying online can be challenging as one feels quite isolated at times. Some of my online tutors really stood out for going the extra mile: Radu-Andrei Nedelcu, Rui Hao, Tanya Reeves and Foaad Haddod deserve special mention for going above and beyond ♥︎. Others were sometimes distant or non-responsive which was quite anxiety-provoking when faced with real difficulties which were holding me back from progressing according to schedule. In a couple of cases I resorted to hiring a local tutor to help me out: it was an added expense but it did mean that I could draw on additional sources of support and avoid falling behind. In addition there is a wonderful network of fellow students going through the program and our WhatsApp groups proved invaluable in establishing friendships and supporting each other. Blake Livermore, Ammar Jamshed, Vaughn Elliot and Hannah Tempest and Philippa Davies are just some of the marvellous people I got to know on the course ♥︎.

Towards the final project…

Many of the modules offer the option to choose your own topic for coursework assignments and I found these were valuable opportunities to pursue areas of special interest to me. For example, in the Python programming module I chose to work on exploring one aspect of operationalizing literary exploration (proposed by Moretti) by automating the identification of characters in novels, their relative importance to the plot, and the sentiments associated with them.

As a result, by the time I reached the point where I needed to choose a final topic for my thesis I had a good idea of the kind of project that I wanted to pursue.

Data science research topics gives you an opportunity to explore different areas of interest, with the aim of producing a working project proposal, which can then be taken through into the the Final project and thesis. While the preceding modules focused on building skills and accumulating knowledge, the emphasis of both research topics and the final project was strongly on academic research: understanding how to conduct a meaningful literature review, how to develop aims and objectives that address a gap in the literature, the importance of choosing an appropriate methodology for your research, and the specifics of academic writing. Coming from a business background I found this all rather arcane and tedious at first! In a business setting one is often given very tight deadline to produce a result and the temptation is to hurry up and just find something that works. What I learned is that this can be shortsighted! The rigours of the academic process brought home to me the value in taking time at the outset to read widely (or at least as widely as possible), before choosing an area to explore more deeply. This not only leads to a better understanding of the area one is working in, but also ensures that one picks a path more likely to yield good results. The discipline of using a well-defined methodology as a framework also highlighted for me how taking a very structured approach can make the ‘next steps’ along the path much clearer.

If you would like to know more about my final project on Automatic knowledge graph construction from news, I made a short video.

Last words

Studying, working full-time, and trying to keep up relationships with family and friends was really tough. The notion of a ‘weekend off’ was lost to me early on! Two years into the program my Dad passed away after a fairly long illness, and simultaneously we were given a week’s extension on our final neural network assignment. I remember sitting in the lounge at 3am watching my neural network train, feeling quite devastated, but at the same time feeling that my Dad would have wanted me to give of my best – even at a time like that! Despite tough times like this, it was worth it! The program has broadened my horizons, honed my skills, and given me a much deeper knowledge on so many interesting areas in the field. It has also left me a strong foundation on which to carry on building… Onwards and upwards 🙂.

PS. I recently found out that during his trial, Nelson Mandela “began studying law with the University of London through distance and flexible learning. In the days before the judge was due to pass down sentencing, Nelson was writing papers for his LLB examination.” Amazing!