Hospital data analytics for business intelligence - An analytics tool for patient feedback analysis.


Student thesis: Doctoral Thesis


Hospitals gather huge amounts of valuable data every day, from clinical studies to patient satisfaction surveys. However, the full utilisation of this data is somewhat lacking with limited data analysis taking place. Improving hospitals data analysis can facilitate the decision-making process leading to informed decisions that will positively influence the patient’s journey. In collaboration with Alder Hey hospital this research identified that the processing time of patient satisfaction data often took 6 to 8 months for analysis. Patients often failed to see any change after making a suggestion or comment, this thesis addresses this issue by utilising machine learning and data visualisation techniques to help reduce the processing time for analysis of text-based patient feedback.
This research focuses on the integration of techniques based on machine learning, neural networks, mathematical modelling(Topological data analysis) and data visualisation to actively assess real-time feedback from Twitter and stored textual data (the Family and Friends Test). To allow greater analysis in a much-reduced time frame. The thesis demonstrates such integration via the creation of a data analytics tool programmed in Python.
Existing approaches require significant computational power to extract the textual information this thesis demonstrates timely extraction with low powered computing systems. Existing solutions do not widely integrate all the techniques used in this thesis, making this research unique by combining the approaches in a low powered computing environment. The research demonstrates the effectiveness of small-scale Machine learning running on low powered computing systems when applied to a text-based extraction and analysis. The research contributes novel algorithms using Topological data analysis and Network theory. The application of this research will aid researchers and business in the utilisation of machine learning and data extraction to support business intelligence gathered from text sources.
Date of Award17 Jul 2020
Original languageEnglish
Awarding Institution
  • Edge Hill University
SupervisorMARCELLO TROVATI (Director of Studies), PETER MATTHEW (Supervisor) & HUAIZHONG ZHANG (Supervisor)


  • Machine learning
  • Hospital Data
  • analytics tool
  • patient feedback analysis

Cite this