Personalized POI Recommendation Based on Subway Network Features and Users’ Historical Behaviors

Danfeng Yan,Zhengkai Guo,Xuan Zhao

doi:10.1155/2018/3698198

Abstract

Current recommender systems often take fusion factors into consideration to realize personalize point-of-interest (POI) recommendation. Historical behavior records and location factors are two kinds of significant features in most of recommendation scenarios. However, existing approaches usually use the Euclidean distance directly without considering the traffic factors. Moreover, the timing characteristics of users’ historical behaviors are not fully utilized. In this paper, we took the restaurant recommendation as an example and proposed a personalized POI recommender system integrating the user profile, restaurant characteristics, users’ historical behavior features, and subway network features. Specifically, the subway network features such as the number of passing stations, waiting time, and transfer times are extracted and a recurrent neural network model is employed to model user behaviors. Experiments were conducted on a real-world dataset and results show that the proposed method significantly outperforms the baselines on two metrics.

Highlights

Recommender system has always been a hot topic and a lot of approaches have been proposed up to now
It significantly outperforms the four other baselines in Mean Absolute Error (MAE) and root mean squared error (RMSE), which demonstrates the effectiveness of our models
W&D-his and W&D-sub outperform W&D, which shows that user historical behaviors and subway network features are important for restaurant recommendation

Summary

Introduction

Recommender system has always been a hot topic and a lot of approaches have been proposed up to now. Different from recommender systems for items like news or music, the location and other geographical factors are more important in personalized POI recommendation This is because the distance between the points of interest and the user mainly determines the travel time, and people usually have more activities like eating, shopping, or watching movies in the region nearby. People usually have to take the transport when they travel at a distance, and public transports such as bus or subway all have their own Wireless Communications and Mobile Computing lines and running schedules In this context, factors such as station location, waiting time, and transfer times have varying degrees of impact on users’ selection and should be taken into account in the POI recommender system. Our work focused on extracting the subway network features and using them to enhance the effect of POI recommendation

Methods

Results

Conclusion