Mining benefits and harms of medical treatments using online health social media
Online health social media allow patients to share their experiences. These patient narratives provide a rich source of information about the benefits and side effects of treatments. Whilst they do not allow quantification of the frequency of such events, they can allow us to identify outcomes and potential issues, and understand the nature of the benefits and harms: the reversibility, the emotional and physical impact, strategies for dealing with the side effect and so on. In the absence of useful information from the medical literature, online social media is a valuable resource for both patients and healthcare professionals, in particular as the size of online social media content increases.
The aims of this project are:
1) to develop a text mining methodology to extract, structure and summarise patient experience from online health social media;
2) to apply the methodology to a specific use case and analyse the benefits and harms for a specific set of treatments.
To address the first aim, the project will explore and develop both rule- and machine-learning based text mining techniques to identify relevant type of content, including themes (e.g. emotional and physical impact) and associated sentiment in patient generated content. This will require mapping of informal and layman terms to medical resources. The information will be modelled using graph databases (e.g. using Neo4j).
For the second aim, we will develop a case study to provide further understanding of the benefits and harms of different treatments for Rheumatoid Arthritis. We will use the data provided by HealthUnlocked, Europe’s largest health social network with almost 200k registered members and a million monthly visitors. Around two million pieces of user-generated content have been created to date. Comparisons will be performed between aggregated graph data as available in HealthUnlocked (the Health Graph) and automatically summarised text mined data.
Student Background: A first class (or equivalent) undergraduate degree in computer science or health informatics, with excellent experience in programming. Interest and some understanding of natural language processing, machine learning and/or social network analytics is an advantage. Interest in health-related research and ability to engage with multi-disciplinary teams are highly desired. Master studies in a related area (including health or bio-informatics) will also be an advantage.
The project will be co-supervised by Dr Goran Nenadic (School of Computer Science, University of Manchester, Health eResearch Centre, HeRC), Dr Will Dixon (Arthritis Research UK & University of Manchester, Health eResearch Centre, HeRC), and Dr Matt Jameson Evans (HealthUnlocked).
More information on current activities in the team: http://gnode1.mib.man.ac.uk/hecta.html
The School has full scholarship opportunities for home and EU students. For international students, the School has fees contribution awards. These awards are awarded on a competitive basis.
Further information on funding can be found here: http://www.cs.manchester.ac.uk/study/postgraduate-research/programmes/phd/funding/
The minimum requirements to get a place in our PhD programme are available from:
How good is research at University of Manchester in Computer Science and Informatics?
FTE Category A staff submitted: 44.86
Research output data provided by the Research Excellence Framework (REF)
Click here to see the results for all UK universities