Public transit is a major mode of urban mobility in Macao, a compact and populous city in China. Public transit ridership is related to round-the-clock shift-based work arrangements in the gaming industry. Smartcards that are used in transit fare collection are a vital data source for understanding the travel patterns and characteristics of transit passengers. However, transactional information from single data sources is fragmented, hindering the extraction of mobility patterns. We proposed a trip-pair-based clustering model that extracts the within-day mobility patterns of bus riders. DBSCAN is employed to filter scattered trips and hierarchical agglomerative clustering to identify clusters. Using a six-month smartcard dataset from Macao, categories of bus passengers were identified. Travel patterns exist with irregular, one-way, and round-trip passengers, and shift workers. The unique shift-based work-life and mixed travel modes in Macao can serve as an example of dense and compact city travel for smart cities.