IoT-Enabled Cross-Field and Reconfigurable Service Provisioning With User-Centered Design.

In the Internet of Things era, people can own or access increasingly interconnected smart and ambient devices. In addition, existing service provisioning (including service recommendation) focuses on single-field (e.g., single store or vendor)-oriented shopping applications, which can only collect fragmented user preference data that are almost always owned by vendors. Furthermore, the current interactions between users’ mobile devices and ambient beacons are unidirectional and noncustomizable, leading to nonreusable and inflexible services. To address the above-mentioned issues, the proposed Market and Discover (MnD) module realizes more user-centered data collection from cross-field sources, and users can own and manage their preference data to enhance privacy. Next, through the Match and Recommend (MnR) module, users can obtain quicker and more efficient service provisioning while automatically reducing unwanted advertising information. With the foundations of MnD and MnR, the proposed Meet and Interact (MnI) module can further help us to establish cross-field, reconfigurable services to realize bidirectionally interactive shopping experiences. In our experimental results, MnD can more comprehensively capture users’ potential cross-field preferences. MnR has made more accurate service provisioning than traditional element-oriented and structure-oriented matching methods, and the execution times for those service-provisioning systems are faster by about 15%–50%. MnR has also helped users filter out unwanted advertisements by approximately 60%. Finally, MnI has enabled users to create their own reconfigurable and reusable services in several testing scenarios. [ABSTRACT FROM AUTHOR]