Date of Award

5-1-2017

Degree Name

Master of Science

Department

Computer Science

First Advisor

Rahimi, Shahram

Abstract

The hybrid correlation extraction method [1] is applied to hospital patients’ satisfaction surveys to provide recommendations for increasing hospital performance and patients’ satisfaction. This method used a multi-layer clustering and labeling system for extracting correlations between features of the patients and hospital. Validating the results of this method is a crucial part of the development, not only for evaluating the proposed method but also to provide a performance measure for each extracted result to the means of comparison, pruning, and ranking. In this thesis, result evaluation methods, their assumptions, and use are studied to find the best method that matches the characteristics of HCAHPS (Hospital Consumer Assessment of Healthcare Providers and Systems) dataset. The methods that fulfill the preliminary assumptions are applied to a simulated HCAHPS dataset with injected correlation. The performance of these methods is analyzed in detail regarding the fact that injected rules are known. The best method that has the best performance and offers comparison threshold for separating strong and weak correlations is selected. This evaluation method is applied to HCAHPS real dataset and the correlations extracted using the hybrid method. The results are used to eliminate irrelevant correlations and rank them from strongest to weakest.

Share

COinS
 

Access

This thesis is only available for download to the SIUC community. Others should
contact the interlibrary loan department of your local library.