Predicting Abandonment in Telehomecare Programs Using Sentiment Analysis: A System Proposal

In the last decades, an increasing attention was posed towards the management and treatment of chronic diseases. This as led to a consequent need of making patients more active in the treatment of their illness (patients empowerment) and to the ever-increasing introduction of telehealth, telemedicine and telehomecare systems. Even though telemedicine services have been proved to improve many aspects of patients’ care, studies have also shown that a relevant percentage of patients abandon telemedicine programs. In this work, a system architecture is proposed in order to monitoring patient’s opinion about a telehomecare service. The presented architecture is composed by three main steps: the development of survey instrument defining a systematic survey’s a survey tool for administrating questions to patients, an analysis module that will perform both Sentiment analysis and Emotion mining on open text answers, and more general Machine Learning techniques to monitor patient’s opinion and make predictions about patients dropout.