Content-Aware Point of Interest Recommendation on Location-Based Social Networks

By Huiji Gao, Jiliang Tang, Xia Hu, Huan Liu. 2015

325 citations recorded by the source corpus.

The rapid urban expansion has greatly extended the physical boundary of users' living area and developed a large number of POIs (points of interest). POI recommendation is a task that facilitates users' urban exploration and helps them filter uninteresting POIs for decision making. While existing work of POI recommendation on location-based social networks (LBSNs) discovers the spatial, temporal, and social patterns of user check-in behavior, the use of content information has not been systematically studied. The various types of content information available on LBSNs could be related to different aspects of a user's check-in action, providing a unique opportunity for POI recommendation. In this work, we study the content information on LBSNs w.r.t. POI properties, user interests, and sentiment indications. We model the three types of information under a unified POI recommendation framework with the consideration of their relationship to check-in actions. The experimental results exhibit the significance of content information in explaining user behavior, and demonstrate its power to improve POI recommendation performance on LBSNs.

Subfields: Geography, Planning and Development, Information Systems, Transportation

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Geography, Planning and Development

1. Jia-Dong Zhang
2. Chi-Yin Chow
3. Jiannong Cao
4. Gao Cong
5. Qiang Yang
6. Yu Zheng

In Information Systems

1. Miao Zhao
2. Tong Chen
3. Xiangyu Zhao
4. Jia-Dong Zhang
5. Chi-Yin Chow
6. Hongwei Wang
7. Jiliang Tang
8. Minyi Guo

In Transportation

1. Jia-Dong Zhang
2. Chi-Yin Chow
3. Jiannong Cao
4. Gao Cong
5. Qiang Yang
6. Jiliang Tang
7. Jun Pang
8. Yang Zhang

Citations in the corpus

• Joint Topic-Semantic-aware Social Recommendation for Online Voting (2017, background): “To mitigate this issue, many approaches have been proposed to utilize social network information in recommender systems [3, 6, 10, 12, 25, 29, 31, 32].”
• Where to Go Next: Modeling Long- and Short-Term User Preferences for Point-of-Interest Recommendation (2020, background): “Early studies in POI recommendation focus mainly on estimating users’ preferences using Collaborative Filtering (CF), especially Matrix Factorization (MF) based techniques (Ye et al. 2011; Cheng et al. 2012; Lian et al. 2014; Gao et al. 2015).”
• Rank-GeoFM: A Ranking based Geographical Factorization Method for Point of Interest Recommendation (2015, background): “Other types of contexts utilized in POI recommendation include category of POI [13], text description of POI [23, 12, 10, 8, 26], and currently-visited POI [4, 6].”
• Instilling Social to Physical: Co-Regularized Heterogeneous Transfer Learning (2016, background): “…of LBSN either tackle link prediction or user profiling on social side with the help of locations (Cho et al. 2011; Scellato and Mascolo 2011), or address location or activity recommendation on physical side with the help of social media (Bhargava et al. 2015; Gao et al. 2015; Zheng et al. 2010).”
• Deep reinforcement learning for page-wise recommendations (2018, background): “Recommender systems have become increasingly popular in recent years, and have been utilized in a variety of domains including movies, music, locations, and social events [2, 11, 13, 30, 31, 39, 41, 45].”
• GeoSoCa: Exploiting Geographical, Social and Categorical Correlations for Point-of-Interest Recommendations (2015, methodology): “To observe the real distribution of the social check-in frequency or rating, we conducted analysis on the two publicly available real-world data sets: Foursquare [6] and Yelp [24].”
• DeepCity: A Feature Learning Framework for Mining Location Check-Ins (2016, background): “In the literature, various choices have been made, such as 10 in [3], 2 in [9], 1,000 in [2], and 40 in [25].”
• Collaborative Filtering-Based Recommendation of Online Social Voting (2017, methodology): “Comparing UGUV + UNN with UNN, UGUV + UVUV with UVUV, UUV(2-hop) + UNN with UNN, and UUV(2-hop) + UVUV with UVUV, we can see that both group information and social information can improve the hit rate of hot votings, while nonhot voting hit rate stays the same or even slightly worse.”