Understanding the emotions of those with a gambling disorder: Insights from automated text analysis

Abstract

The diffusion and growth of the web and the social media applications that it has provided have seen people increasingly turn to social media to express their feelings, frustrations, and ambitions and to generally share life events. Like other internet users, those with a gambling disorder are also known to use specialized online forums to read the experiences articulated by others and to open up, share, and express themselves online. In this study, we examined how those with a gambling disorder talk about their emotions and express sentiment through their online comments. Sentiment Analysis in IBM Watson was used to capture expressed emotions (anger, disgust, fear, happiness, sadness, and surprise) and sentiments. These data were then used as input to a latent class cluster modelling procedure aimed at categorizing those with a gambling disorder into distinct groups. The findings show how qualitative online data can be transformed into quantitative insights in order to identify different categories of people with a gambling disorder. The technique offers a non-intrusive method of data collection that can provide useful insights into the emotions felt and expressed by those with a gambling disorder.