PurposeQuality 4.0 is a new paradigm of quality management, which emphasises the need to adapt to recent technological innovations by updating traditional quality approaches. Amongst the most important factors for adopting Quality 4.0 is the leveraging of big data to collect insights and quality perceptions from clients. Therefore, user reviews have emerged as a valuable source of information, which can be analysed through machine learning procedures to uncover latent quality dimensions.Design/methodology/approachThis study applies a combination of text mining techniques to analyse Airbnb reviews, identifying service quality attributes and assessing their relation to the users' sentiment. More than two million reviews written by guests in four European cities are analysed. First, topic modelling is applied to find the quality attributes mentioned by reviewers. Then, sentiment analysis is used to assess the positiveness/negativeness of the users' feedback.FindingsA total of 37 quality attributes are identified. Four of them show a significant positive relation to the guest's sentiment: apartment views, host tips and advice, location and host friendliness. On the other hand, the following attributes are negatively correlated with user sentiment: sleep disturbance, website responsiveness, thermal management and hygiene issues.Originality/valueThis paper provides a practical example of how Quality 4.0 can be implemented, proposing a data-driven methodology to extract service quality attributes from user-generated content. Additionally, several attributes that had not appeared in existing Airbnb studies are identified, which can serve as a reference to extend previous quality assessment scales.