University of Twente Student Theses

Login

Analysis of unstructured data at GGD : text mining semi-medical notes to predict childhood obesity and abuse

Paauw, Tim (2015) Analysis of unstructured data at GGD : text mining semi-medical notes to predict childhood obesity and abuse.

[img] PDF
Restricted to Restricted

3MB
Abstract:Abuse and obesity are two important threats to children’s health that the Jeugd Gezondheidszorg (JGZ) in Amsterdam tries to prevent. Analysis of data is a part of this effort. This data, registered by pediatricians during consultations with children, is partly structured and partly unstructured. The unstructured part consists of semi-medical notes written by pediatricians, containing all sorts of information on the well-being of a child. This research at Big Fellows explores methods of analyzing the unstructured data with respect to abuse and obesity. For obesity this means being able to predict whether a child will suffer from obesity at a future point in time. Abuse is often not registered and the goal here is to aid detection and registration by indicating the chances of a child being abused. Exploratory research is done to get insight and to assess the feasibility of predictive models. It becomes clear that the data can be used for detecting presumed abuse, but for obesity other structured data gives better results. Focusing on detecting abuse, predictive models are built and compared. The model is then wrapped in an API which is implemented in the decision support tool that the JGZ already uses.
Item Type:Essay (Master)
Clients:
Big Fellows, Utrecht, Netherlands
GGD Amsterdam, Amsterdam, Netherlands
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:54 computer science, 88 social and public administration
Programme:Business Information Technology MSc (60025)
Link to this item:http://purl.utwente.nl/essays/67797
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page