Abstract: 【Purpose/significance】Online user reviews are feedback on the user's perception of the experience of a product or service.Text mining of online comments is an important part of intelligence analysis.【Method/process】In order to mine users’interest from massive online comments and provide guidance for website manager to strengthen management, this paper crawls online reviews of ho⁃tel users in four cities on TripAdvisor website and uses text clustering algorithm to construct the clustering map based on the theory of topic map. Map visualization reveals the differences of hotel users' views in different regions. Finally, the features of social networks and the centrality of keywords in different maps are analyzed.【Result/conclusion】This study found that hotel users are most con⁃
cerned about the service, followed by the hotel environment and location. The research can quickly explore hotel users' concerns, help hotel managers understand the needs of users for accommodation and in consequence improve user satisfaction.【Innovation/limitation】This paper combines topic map and text mining technology to construct the topic map of hotel users’online reviews, which shows supe⁃riority in big data text topic clustering. However, this paper only analyzes the online reviews of some hotels in four cities on TripAdvi⁃sor, therefore the data coverage is not extensive enough.