Importance-Performance Analysis Based on Text Mining

Purpose: Importance-performance analysis (IPA) has gained significant recognition in research of service quality. This study enhances the existing body of knowledge in three significant ways. Firstly, the primary aim of this article is to demonstrate how to conduct IPA using text mining techniques to analyze opinions scraped from the internet, eliminating the need for extensive questionnaires (novel approach). Secondly, we propose a simple statistical data adjustment technique based on the Cronbach alpha coefficient. Thirdly, we provide a literature survey of the evolution of IPA in time. Design/Methodology/Approach: The narrative review of literature subject was used as well as three case studies applying novel approach to IPA in examples of three different hotels in Poland. Findings: We assessed the effectiveness of the novel approach through case studies of three hotels of varying quality. In our opinion, all hotels were accurately diagnosed in the IPA conducted. Practical implications: We believe our approach is flexible enough to accommodate further enhancements, including both the techniques for extracting information from text and refining the IPA itself. Orginality/Value: IPA based on text mining is an interesting alternative to the traditional approach which typically involves researching hundreds or even thousands of respondents through personal questioning.