Word Cloud Analysis of Customer Satisfaction in Cosmetic Products in Thailand

Abstract

This research aims to investigate customer satisfaction in cosmetic products by utilizing a low-cost tool, word cloud, to analyze online reviews, based on the research questions: how customers feel about each type of cosmetic products? what their feedbacks are? and which words are positive and which ones are negative? The dataset for the investigation comprises with reviews over a 4-year duration of data collection from 2015-2018, collected from popular social networking sites in Thailand including Facebook and Pantip. A hierarchical clustering approach, the Linkage algorithm, was employed in the context of text mining. The result shows that factors that influence customer satisfaction are based on customer experience affecting positive or negative words in online reviews of each product.