Many of us are familiar with recommender systems, especially from sales platforms. Based on our previous decisions, they recommend other products that could also be of interest to us. But can they guide us towards “better” options that are genuinely more beneficial for us? Mathias Jesse, doctoral student in the doctoral school DECIDE, is investigating the working mechanisms of various technical concepts.
Let’s imagine that we are browsing a recipe platform for Christmas biscuit recipes that are usually full of sugar and fat. Nowadays, it is widely known that the ingredients in biscuits are not necessarily beneficial for human health. In his thesis, Mathias Jesse will be examining which “enticements” can be used to steer us towards healthier, low-sugar and low-fat biscuit recipes by means of nudging.
His research revolves around the so-called recommender systems. Based on previous decisions, these systems recommend other things that we might like or might be interested in. While sales platforms primarily want to boost their profits, there are, as Mathias Jesse explains, some applications with more lofty goals, as the example of the recipe platform shows.
We ask him whether there is, in fact, a fair amount of manipulation involved in influencing people to come to “better” decisions. Mathias Jesse tells us: “People are not always able to make the best possible decision for themselves. Sometimes they need a little push, a so-called nudge.” The term originally comes from behavioural economics and was first introduced by Richard Thaler and Cass Sunstein. They see a nudge as a method of persuading people to behave in a way that is better for them, without applying prohibitions and rules. So, if someone is hunting for Christmas biscuit recipes, they are more likely to be shown those recipes that have fewer detrimental effects on our blood sugar and cholesterol levels or our hip and stomach measurements.
Mathias Jesse programmes various kinds of applications for this type of technical nudging and investigates the extent to which these also produce the desired effects. One example is the “default” function. If the healthier alternative is predefined as the default case, people will be more likely to choose it, so the assumption goes. A multitude of psychological, economic, but also ethical factors underpin the different applications, which, as Mathias Jesse explains, need to be considered from an interdisciplinary perspective.
Ever since his school days at the HTL Mössingerstraße, Mathias Jesse has been interested in the intersections between computer science and other subjects. “Computer science alone seemed too boring”, he tells us. In the end, the Information Management degree was exactly what he was looking for, combining the best of two worlds: technical knowledge from computer science and business management expertise from business administration studies. Following the completion of his Master’s degree programme, Mathias Jesse applied for a place in the doctoral school DECIDE (Decision-making in a digital environment). The topic of his thesis resulted from discussions with his supervisor Dietmar Jannach. Mathias Jesse adds: “For quite some time now, the interaction between man and machine, i.e. including the user interfaces on websites, has been my main focus. So, nudging in conjunction with recommendation systems seemed the ideal solution.”