A Text Mining Approach to Measuring & Predicting Perceived Service Quality from Online Chatter

Online reviews have been studied for their potential to explain consumer behaviour and to predict business performance. However, large volume and unstructured format of the textual part of the reviews make them unamenable to analysis with traditional methods like survey. By employing the automated text mining methodology this study explores how consumers reflect on the perceived service quality in the online reviews. Specifically, sentiment analysis, text classification and predictive modelling will be applied to investigate the dimensions of perceived service quality together with their contribution to the overall rating of service experience. The results are expected to extend the understanding of perceived service quality; to illustrate the value of the automated text mining methods for maximising the usefulness of online reviews and helping businesses reach their strategic goals.