PRMF: point of interest recommendation method integrating multiple factors
Ting Yu, Lihua Zhang, Yinhao Zhang · International Journal of Web Engineering and Technology · 2023
Point-of-interest (POI) recommendation plays an increasingly important role in location-based social networks (LBSNs) and is widely used in various e-commerce websites. However, due to the high sparsity of user check-in information, it is still challenging to recommend appropriate and accurate locations to users. As people decide where to visit based on numerous factors, recommendation systems need to consider check-in records and data on POI popularity and POI locations. In this paper, we propose a POI recommendation method that integrates multiple factors by analysing users' check-in records, POI category, location, and POI popularity, called PRMF. Firstly, we employ a neural network algorithm to calculate user preferences. Activity centres are then calculated based on the users' historical check-in history, and geographical preferences for each POI are calculated according to the activity centre. By combining the popularity of POIs in this study, we calculate POI popularity preferences, and the above three parts were obtained by linear fusion to calculate the users' final preference. Extensive experiments (based on real datasets, including long-term check-in data for locations in New York and Tokyo collected from Foursquare) show that our proposed method was superior to the baselines.