A collaborative filtering approach for mobile commerce applications to push notifications on mobile devices and wearable accessories

Mobile devices have been a major driving force for electronic commerce since the emergence of smartphones. As smartphone users embrace m-commerce applications, vendors can push personalised advertising and promotion information anytime and anywhere to their smartphones in the form of notifications. If notifying through the smartphone is not enough to guarantee our attention, smartwatches and other smartphone-linked wearable accessories can help grab our attention with even more accessible notifications. Some smartphone users who get too many notifications may opt to restrict or filter out unimportant notifications and there are tools readily available for that. Yet, there are cases where notifications are welcome but they simply arrive at inconvenient times or in inappropriate manners, e.g. audible alerts in a museum. It is therefore desirable for m-commerce applications to be intelligent in delivering notifications at appropriate times and in appropriate forms based on user preferences. While there exists algorithms that train an app to deliver notifications appropriately on an individual basis, a mass-market m-commerce smartphone application needs to take into account the diverse profiles of vast user population. This research addresses the notification delivery problem based on a collaborative filtering approach that involves collecting in situ user feedback on notifications.