Identifying User Experiences for Decision-Making in Service Science

The purpose of this paper is to propose a method based on text mining techniques that employs user comments on the Web to clarify users’ experiences of a service. The method classifies user experiences into the categories of instrumental qualities, non-instrumental qualities, and emotional reactions. The Text-Miner Software Kit (TMSK) and Text Rule Induction Kit (RIKTEXT) were used to obtain the classification rule sets. A case study on the tourist domain was conducted to assess the rules’ precision in classifying sentences into the three user experience categories. One thousand hotel reviews from the website www.​tripadvisor.​com were used in the evaluation process. Two datasets were built: one for training and one for testing. For validation, 100 new opinions were used to assess the rules’ precision. The results are considered good because there was only a 10% error in automatic classification versus manual classification. Based on this result, we argue that the automatic identification of information about user experience of a service can be done accurately using text mining techniques, such as those presented in this work.