I work as a Researcher at the University of Twente since December 2013. Until 2016 I combined my PhD research with a part-time function at IGS Datalab, for which I developed policy and advised on research data management and analysis. My PhD research focuses on developing new methods and techniques for analyzing complex, unstructured data sets (e.g. written or spoken text). I use text mining, natural language processing, machine learning, audio signal processing, and data visualization to develop supervised classification models. Such models can be used to screen for mental health disorders (such as PTSD or depression), to monitor treatment progress, or to predict treatment outcomes. I have also contributed to a pilot study on fraud detection in social benefits, commissioned by the Dutch Ministry of Social Affairs and Employment. My research is supervised by Prof. Bernard Veldkamp (UT) and Prof. Miranda Olff (AMC).
Education & Previous experience
I completed my Master Educational Science & Technology in 2011. My graduate research focused on the development and validation of complex IQ test items to differentiate on higher levels of intelligence. After graduating I worked as a Data Analyst at the department of Research Methodology, Measurement and Data Analysis and the Medisch Spectrum Twente (MST) for half a year. In this position I contributed to a study on sleep apnea by analyzing and validating several screening instruments using Item Response Theory.
My research focuses on the application of modern methods and techniques like machine learning, text mining, natural language processing, audio signal processing, and data visualization to unstructured data sets (like written or spoken text) from the mental health domain.
Wiegersma, S., Mink-Nijdam, M.J., Van Hessen, A.J., Olff, M., & Veldkamp, B.P. (2017, June). Recognizing hotspots in Brief Eclectic Psychotherapy for PTSD by text and audio mining. Presentation at the 15th European Society for Traumatic Stress Studies conference, Odense, Denmark.
Wiegersma, S., Van Noije, A.J., Sools, A.M., & Veldkamp, B.P. (2016, October). DIY Text Classification – The development of a supervised text classification tool. Presentation at the RCEC workshop Item Response Theory and Educational Measurement, Enschede.
Wiegersma, S., Sools, A.M., & Veldkamp, B.P. (2016, July). Exploring “Letters from the Future” by visualizing narrative structure. Presentation at the Computational Models of Narrative conference, satellite symposium van de Digital Humanities conference, Krakow, Poland.
Wiegersma, S., Sools, A.M., Veldkamp, B.P., & Westerhof, G.J. (2016, May). What works when for whom? Advancing therapy change process research by mining for therapy-related textual features in effective e-mental health interventions. Presentation at the conference Supporting Health by Technology VII, Groningen.
Wiegersma, S. (2015, June). IGS Datalab - Research Data Support. Poster presentation at the Netherlands eScience Center Roadshow, Enschede.
Wiegersma, S., & Van Tongeren, S.J. (2015, June). Research Data Management. Presentation at the IGS PhD Day, Enschede.
Wiegersma, S., & He, Q. (2014, July). Screening for PTSD using verbal features in self-narratives. A text mining approach. Presentation at the Summer School Big Data in Clinical Medicine, Enschede.
Wiegersma, S., He, Q., & Veldkamp, B.P. (2014, May). Screening for PTSD: A text mining approach. Poster presentation at the international symposium The impact of great wars and beyond, Leiden.