Exponentially weighted moving average control charts for monitoring customer service quality comments
Triss Ashton, Nicholas E. Evangelopoulos, Victor R. Prybutok · International Journal of Services and Standards · 2013
This paper examines the use of variable control charts with data that originates in text form and uses factors extracted from the text by latent semantic analysis. We demonstrate how text data from customer comments is analysed as well as the steps necessary to visualise the extracted factors on an Exponentially Weighted Moving Average (EWMA) chart. We also show how the factors correspond to latent service related issues and how using the EWMA chart to monitor those factors allows for addressing and improving service quality issues. We also review issues in text mining and suggest areas for future research.