Personalised News and Blog Recommendations based on User Location, Facebook and Twitter User Profiling

Abstract

This demo presents a prototype mobile app that provides out-of-the-box personalised content recommendations to its users by leveraging and combining the user's location, their Facebook and/or Twitter feed and their in-app actions to automatically infer their interests. We build individual models for each user and each location. At retrieval time we construct the user's personalised feed by mixing different sources of content-based recommendations with content directly from their Facebook/Twitter feeds, locally trending articles and content propagated through their in-app social network. Both explicit and implicit feedback signals from the users' interactions with their recommendations are used to update their interests models and to learn their preferences over the different content sources.