Real-time smartphone sensing and recommendations towards context-awareness shopping

This study investigates a smart environment, namely “Intelligent Shopping-aid Sensing System (iS³)” for online shopping support in the next era by developing a context-aware automated service system. Sensors, radio frequency identification (RFID), are applied for the recognition, collection, and delivery of user contexts. Following the collected contexts from sensors, integrated mining and analysis techniques (i.e., customized clustering analysis and association rules) were implemented for the provision of instant and personal information to users. Information of products, such as locations, specifications, and characteristics can be collected quickly through the deployed RFID reader and display. Moreover, local applications on mobile devices offer real-time interactions between central system and end users. The system is expected to prompt the product promotion, inquiry and online marketing to shopping malls (and related companies as well). In the empirical results, the quality of recommendations with the proposed approach reaches 70 % accuracy rate. The traditional and non-clustering approaches are 56 and 46 %, respectively. This study reduces long-term operation costs of retailers, stimulates service innovation and experience economy and enhances corporate operational performance.