A user-generated content-based social network large-scale group decision-making approach in healthcare service: Case study of general practitioners selection in UK

Globally rising concern of healthcare has highlighted increasing demands for healthcare management. With the web-based social media and platform spreading in the healthcare industry, online reviews of patients have emerged as a crucial indicator and effective instrument for evaluating the service of healthcare providers. Healthcare management involved by many stakeholders could be considered as a complex and mutli-faceted group decision making scenarios. Therefore, through mining patients’ concern from the user-generated content (UGC), this study aims to develop a comprehensive large-scale group decision making (LSGDM) approach for healthcare service management. Firstly, by utilizing topic modelling techniques, various criteria with respect to healthcare service are retrieved from online reviews and criterion weights are also derived. Secondly, a novel clustering method based on node potential influence and the m-ary adjacency relation is designed to categorize large-scale decision makers (DMs) into small and manageable clusters. Thirdly, considering that the loss of evaluation information occurs during decision making process, Cornish-Fisher expansion is utilized to deduce the mean, standard variance, and kurtosis estimators of the incomplete information, which are further converted into cloud models by the proposed backward cloud transformation algorithm. Finally, a minimizing bi-capacity entropy optimization model is constructed to derive 2-additive bi-capacity parameters that are adopted to depict the interactions between clusters. A bipolar Choquet integral information aggregation approach is then presented to aggregate preferences of different clusters to rank alternatives. A case study on General Practitioners medical service in United Kingdom and a comparative analysis are further performed to validate our proposal.