Presence: Mo, tue, wed, fri
Room: Cubicus B126b
Phone number: +31 53 489 4564; +31 53 489 4470 (secr.)
Multivariate statistics, missing data, causal inference, text mining, natural language processing.
I am a PhD student working on the What Works When for Whom project. The project is a collaboration between the University of Twente and the e-Science center in Amsterdam. The supervisors for my project are Anneke Sools, Sytske Wiegersma and Janneke van der Zwaan, and the project is promoted by Gerben Westerhof and Bernard Veldkamp.
The project is focused in the area of eHealth, which encompasses the emails that clients exchange with their therapist. For some clients, this provides noticeably improved results in their therapy, while for others the effect is much smaller. From where does this difference arise? The aim the project is to examine e-mail contact between clients and their therapist for indications as to the reasons for the effect. What is communicated in the emails? What is indicated about the relationship between the client and their therapist? How does the relationship develop over time?
These questions are – of course – qualitative research questions, but with more than 10.000 e-mails, such approaches are unfeasible. Thus an automated approach is preferred. To quantify qualitative methods, techniques such as text mining, natural language processing and other machine learning techniques will be utilized to address the Big Data issue. The complexities of the overall problem necessitates a multi-disciplinary approach. As a statistician, I am enthused by the opportunity this provides for me to make a significant and meaningful contribution to the health of patients.