Public Graduation Presentations, January 26
On Friday January 26 we organize the Public Graduation Presentations of three Media Technology MSc graduates. In 20-25 minutes, each student presents their Graduation Research Project, followed by 10 minutes public discussion.
Everyone is welcome to attend!
When and where?
Friday, January 26, 2018.
Room 0.11 of the Gravensteen building,
Pieterskerkhof 6, Leiden (downtown).
15:30 - 16:05 Georgios Bouzias
16:15 - 16:50 Stijn van Linden
17:00 - 17:35 Sander de Bont
15.30 - 16.05 Georgios Bouzias
Title: “User verification by air-drawing”
Abstract: The aim of this project is to build and evaluate a verification system on air-drawings of a simple symbol. The air-drawings data were collected with a prototype specifically made for this purpose that utilizes the depth camera of a Kinect v2 in an experimental setting. Its purpose is to capture the drawing trace a user can make with an extended index finger. The classifier used is a one-class SVM, and its evaluation is made with precision and recall metrics. In addition, a distinction is made between users with different familiarity with the system. The intent is to study separately a group of users that have only a basic familiarity with the system, and a very experienced user. Conclusively, we would say that given the performance, a verification system as discussed in this study, could possibly be used only in issues that are less sensitive than locking or unlocking a door, which is the main scenario discussed in the study.
Thesis adviser: Edwin van der Heide and Peter van der Putten
16.15 - 16.50 Stijn van Linden
Title: “Personality as a factor for knowledge sharing behaviour in novices”
Abstract: Knowledge sharing is at the core of any good organisation or team, and has become an elaborate field of study (Alavi & Leidner, 2001). Moser & Wodzicki (2017) studied under what condition experts shared their knowledge in groups. In this study we turn it around and investigate if laypeople dare to share a different opinion than an expert in a student environment, and if their knowledge sharing behaviour can be predicted by personality traits. The study is conducted through a survey, containing a personality questionnaire and a thought experiment. Participants were split in two groups (1) an experimental condition that placed participants with an expert and (2) a control condition that placed participants with peers. Chi-square test was used to check for significant differences but none were found p = .08. Distribution was heavily in favour of sharing knowledge contradicting the experts opinion. The only significant predictor found for not going against the expert was the personality trait Agreeableness p = .045.
Thesis advisors: Max van Duijn and Erik de Kwaadsteniet
17.00 - 17.35 Sander de Bont
Title: ”Online Persuasive Learning”
Abstract: Robert Cialdini’s six principles of persuasion are being applied with success in sales for decades. A lot of commercially driven organisations continue to discover new implementations of the key principles on how to influence human behaviour online. And, as a result, increase the knowledge about online persuasion. Which results in a greater return in the form of time spent or conversion rate. Online learning environments, such as Massive Online Open Courses (MOOCs), are known for their high dropout rates of MOOC learners. MOOC platforms could also benefit from these new implementations. And, moreover, the MOOC learners could make the most out of the course and their learning intention. One of persuasion principles is social proof. This principle is often applied with success in online sales. Therefore, this research aims to investigate the effect of a specific social proof implementation in a MOOC course. The course is hosted on the Coursera platform, and the experimental setup is tested with an A/B test. On three critical moments in a MOOC, the learner is exposed to a simple notification about other learners behaviour during those moments. A positive effect of this social proof implementation is expected on the completion rates.
Thesis advisors: Bas Haring and Tanja de Bie