Opinion analysis and statistics about the success rates of treatments extracted from online healthcare forums for the disease psoriasis using a pipeline of LDA, NLP and artificial neural network

Abstract

This paper describes how the authors have applied NLP and text processing techniques to mine the messages related to the disease Psoriasis available online and posted by users who have undergone various types of treatments for the disease. The authors discuss the various classification techniques applied to classify the extracted messages as solution for the disease or non-solution. The authors also provide statistics regarding the success rates of the various types of treatments and medications in a consolidated form. A comparison study of the different types of treatments and their success rates is performed and statistics provided to the end users.