Next point of interest (POI) recommendation

Point-of-interest (POI) recommendation is an imperative benefit to Location-Based Social Systems (LBSNs) that can advantage both users and businesses. POI recommendation can help public to discover new interesting locations. The user and location data are also easily available these days due to the location-based social networks that enable users to share their check-in information through applications. POI recommendation is upcoming, and many research techniques have been suggested and investigated in this field over time. In this project, some existing next POI recommendation algorithms are explored and developed further on a public dataset obtained from open source applications. This project also involves the evaluation of some common recommender system techniques, applied to POI check-ins. From the evaluation, we get a few important discoveries in order to utilize POI recommendation models in different scenarios. Moreover, we observed the trends in Singapore data. It was discovered that many POIs present in the data have a combined infrastructure location space. This meant that those POIs shared the same address. Certainly, it was true to Singapore city due to its smaller land, many outlets are in the same infrastructure. Thus, we propose a method to group such POIs into combined spaces and carry out POI recommendation techniques on the combined POI data and compare the results accordingly as several new or interesting POIs in the same location point to be recommended for users can be a more meaningful way of recommendation.Bachelor of Engineering (Computer Science