LAUSR

Online review platforms offer customers the opportunities to express valuable feedback and personal views from various aspects on restaurants, products, works of art, or other items. A majority of previous studies on these user-generated reviews are devoted to controversy, bias, and opinion analysis. However, little work has been done to study urban characteristics via topic analysis from the city level in reviews. In this paper, we propose a visual analytics system, to visually explore spatiotemporal urban topics for cultural trend discovery, location mining, and decision making. Specifying a topic by users is supported due to the difference between review text and traditional text, such as news and books, and the diversity of topics and users. Sentiment analysis and statistical analysis are adopted to characterize the temporal trend and sentiment and topic geographical distributions of the user-specific topic. Our system allows the user to interactively explore the time-evolving frequency trend and characteristic geographical distributions of a topic in reviews. We evaluate the effectiveness and usefulness of our system using three case studies in different domains.Graphical Abstract