William Goldsworthy Undergraduate Dissertation 2017/18

Medical study identification for the ScHARRHUD database using text mining techniques

Supervised by M.Stevenson

Abstract

There is currently a large and growing number of published medical studies which makes the identification of relevant studies for systematic reviews increasingly difficult and time consuming. Experts are forced to manually sift through thousands of irrelevant studies in order to identify those that are relevant and useful for their review. ScHARRHUD is an innovative database containing bibliographic details of medical studies which report on the health state utility values (HSUVs). This project aims to use text mining techniques to identify studies similar to those already included in the database and consequently improve the quality of the collection. This project will produce a list of relevant studies that can be added to the ScHARRHUD database.