Dr Jacqueline Kazmaier one of the most capable, creative and dedicated postgraduate students, says supervisor
Three years after obtaining her BEng degree and the Chancellor’s Medal, the highest honour bestowed on a student by Stellenbosch University, Jacqueline Kazmaier will receive her PhD in December 2020.
Prof Jan van Vuuren, her supervisor says: “Jacqueline was a truly exceptional doctoral student. Indeed, she was one of the most capable, creative and dedicated postgraduate students whom I had the privilege to supervise during my career of 26 years. She tackled a very difficult problem for her dissertation, but her tenacity and her meticulous approach to research enabled her to make a resounding success of her studies. She was an inspiration – not only to me, but also to all the members of my research group, the Stellenbosch Unit for Operations Research in Engineering (SUnORE), as well as the industry partner attached to her dissertation.
Dr Kazmaier explains why she studied Engineering, and Industrial Engineering in particular: “I have always had an affinity for mathematics and problem solving – the Engineering field attracted me because it is centred around a structured approach to solving problems, using applied maths to drive decision making. In addition to my love of problem solving, I have always had an interest in business. Industrial Engineering applies the problem-solving techniques of Engineering at a systems level, making it perfectly suited to optimising the way that organisations of all kinds function. The classmates that studied with me have gone into vastly different industries, including the automotive, retail, brewing, banking, management consulting, technology and healthcare industries. This illustrates how broad the field of Industrial Engineering (IE) is, and the vast opportunities it provides for a future career. On a different note, the IE department at SU is like a big family. It’s much easier to succeed in your studies and love what you do when the environment you’re in is supportive, motivating and fun – this is something the Department has done extremely well, and this significantly motivated my choice.”
She originally hails from Windhoek, Namibia, and has German ancestry. She says: “I went to an international school, where I was primarily exposed to European culture. When I started university, I wanted to experience South African varsity culture. I was particularly interested in Stellenbosch University because of the ‘school spirit’ and feeling of unity evident in, for example, the Varsity Cup. Living on campus was also a major drawcard, as it further intensifies the familial feeling among students. Finally, the personal response and support I received from the University after I applied far exceeded my expectations and were unmatched by any other university I applied to.”
She completed her final-year project in the SUnORE research group. “I decided after that year to continue with postgraduate studies for two main reasons. First, I had only just scratched the surface of the research area of machine learning and I felt that a postgraduate degree provided the ideal conditions to learn more about the field, and to possibly find a specialty that I’d be interested in. I had grown very passionate about pursuing one of the many opportunities in the machine learning/AI space as a career and was determined to expand my skill set in this direction. Secondly, after having been exposed to the stimulating and supportive environment that the SUnORE group provides, I was inspired to continue my studies and to further learn from the senior students and lecturers in the group. I found that many of the skills I acquired actually related to interpersonal skills, networking and presentation, which have been invaluable.
“I met my supervisor, Prof Jan van Vuuren, during our third year Operations Research course, and was immediately inspired by his meticulous approach to teaching, his passion for his work and his selfless devotion to helping every one of his students achieve their full academic and personal potential. I was thrilled to be able to do my master’s and subsequent PhD under his guidance. In the last five years, Prof has opened my mind to a vast range of new ideas, and shaped the way that I think about and approach many things. I will always be grateful to him for that. My topic was initially a project proposal from an industry partner in the banking industry, but it took shape and moved slightly further away from the initial concept as I explored the literature.”
Her topic was A framework for evaluating unstructured text data using sentiment analysis. On this, she elaborates: “Unstructured data, such as customer reviews, chat forums and news articles, have become a more and more prominent source of information over the last few years. Many of these data sources also contain some degree of sentiment (e.g. a message from a complaining customer, a news article satirising a politician). While it’s relatively easy for humans to read and make sense of such data, automating these tasks presents considerable challenges. The focus of my research was creating a framework that can be used to automatically extract actionable insights from a large collection of opinion-bearing text documents to drive decision making. In one of my case studies, for example, the framework was applied to determine the issues that were raised in customer complaints about a retail bank, and, moreover, to link these issues to specific customer or branch profiles.”
Although the Covid-19 pandemic had some drawbacks for her, it did not affect her research. She notes: “It was a little sad not to be able to work with the other members of my research group in the lab. I definitely missed physical interactions during tea time and the casual conversations about each other’s work. I was very lucky, however, in that we were all able to take our equipment home and work from there. Fortunately, I didn’t have to do any physical experiments and was able to complete everything remotely without delay.
“After handing in my PhD, I started working for Cape AI as a Data Scientist. The company is a tech start-up with a mission to create value in a sustainable way. I’m focusing on projects in the natural language processing domain at the moment, which ties in very nicely with my dissertation topic. So far, I’ve been involved in a project in the banking sector, an internal venture to enhance collaboration between members of an organisation and an exciting partnership in the healthcare industry.”
Topic: A framework for evaluating unstructured text data using sentiment analysis.
The study of opinion has become a necessity in modern industry. Due to the vastness of such data, however, manual approaches towards achieving this are no longer viable. This situation has given rise to the field of sentiment analysis – the computational study of people’s opinions, attitudes and emotions. In this dissertation, a generic framework for sentiment analysis is proposed, facilitating a robust model development process and a deep, versatile analysis of model results by utilising a data mining approach. The framework can aid organisations in successfully leveraging unstructured, opinion-bearing data in combination with structured data sources to inform decision making.
Dr Jacqueline Kazmaier (right) and her supervisor, Prof Jan van Vuuren.