Fused Matrix Factorization with Geographical and Social Influence in Location-Based Social Networks

By Chen-Kuang Cheng, Haiqin Yang, Irwin King, Michael R. Lyu. AAAI Conference on Artificial Intelligence, 2012

643 citations recorded by the source corpus.

Recently, location-based social networks (LBSNs), such as Gowalla, Foursquare, Facebook, and Brightkite, etc., have attracted millions of users to share their social friendship and their locations via check-ins. The available check-in information makes it possible to mine users’ preference on locations and to provide favorite recommendations. Personalized Point-of-interest (POI) recommendation is a significant task in LBSNs since it can help targeted users explore their surroundings as well as help third-party developers to provide personalized services. To solve this task, matrix factorization is a promising tool due to its success in recommender systems. However, previously proposed matrix factorization (MF) methods do not explore geographical influence, e.g., multi-center check-in property, which yields suboptimal solutions for the recommendation. In this paper, to the best of our knowledge, we are the first to fuse MF with geographical and social influence for POI recommendation in LBSNs. We first capture the geographical influence via modeling the probability of a user’s check-in on a location as a Multi-center Gaussian Model (MGM). Next, we include social information and fuse the geographical influence into a generalized matrix factorization framework. Our solution to POI recommendation is efficient and scales linearly with the number of observations. Finally, we conduct thorough experiments on a large-scale real-world LBSNs dataset and demonstrate that the fused matrix factorization framework with MGM utilizes the distance information sufficiently and outperforms other state-of-the-art methods significantly.

Subfields: Electrical and Electronic Engineering, Information Systems, Transportation

View Paper: https://www.semanticscholar.org/paper/1b233b82ba35c913e0ddcf19b11ec50d013a8149

How researchers see this paper

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

In Electrical and Electronic Engineering

1. Liang Wang
2. Tien-Ping Tan
3. Xing Xie
4. Gao Cong
5. Quan Yuan
6. Chunyan Miao
7. Xia Hu
8. Enhong Chen

In Information Systems

1. Xia Hu
2. Huan Liu
3. Jiliang Tang
4. Xing Xie
5. Fabio A. Crestani
6. Hongzhi Yin
7. Huiji Gao
8. Aixin Sun

In Transportation

1. Yu Zheng
2. Xing Xie
3. Aixin Sun
4. Guangzhong Sun
5. Hui Xiong
6. Gao Cong
7. Yanjie Fu
8. Zijun Yao

Citations in the corpus

• Where You Like to Go Next: Successive Point-of-Interest Recommendation (2013, neutral): “The other line of work focuses on LBSN data, which is very sparse and large-scale [Ye et al., 2010; 2011; Cheng et al., 2012].”
• Personalized Ranking Metric Embedding for Next New POI Recommendation (2015, methodology): “Another CF method is Matrix Factorization [Cheng et al., 2012; Lian et al., 2014], which learns the general taste of a user by factorizing the observed user-item preference matrix.”
• Combining Heterogenous Social and Geographical Information for Event Recommendation (2014, background): “Recommendation with geographical information has been extensively studied in recent years (Cheng et al. 2012; Ye et al. 2011; Zheng et al. 2010; Qiao et al. 2014).”
• Practical Privacy Preserving POI Recommendation (2020, methodology): “Therefore, we build the user adjacent network by using user geographical information, similar as the existing researches [13, 63].”
• User Preference Learning with Multiple Information Fusion for Restaurant Recommendation (2014, background): “4 Modeling Geographic Proximity We argue that the nearer the restaurant is from the user’s location, the more likely the user will visit this restaurant [22, 7].”
• GeoMF: joint geographical modeling and matrix factorization for point-of-interest recommendation (2014, methodology): “Concentrating on modeling the distance distribution may ignore the multi-center characteristics of individual visiting locations according to [2].”
• Content-Aware Point of Interest Recommendation on Location-Based Social Networks (2015, background): “Cheng (Cheng et al. 2012) investigated the geographical and social influence through a multi-center Gaussian model.”
• Learning Graph-based POI Embedding for Location-based Recommendation (2016, background): “Specifically, to overcome the data sparsity issue, most prior work of location-based recommendation focused on exploiting the geographical influence [10, 3, 20] and temporal cyclic effect [7, 28] to provide spatial or/and temporal context-aware recommendation.”
• Time-aware point-of-interest recommendation (2013, background): “There exists other work that incorporates social link information into POI recommendations, such as the probabilistic generative model-based method [21], and matrix factorization-based method [4].”
• Learning geographical preferences for point-of-interest recommendation (2013, background): “[6] considered the geographical influence, the multi-center of user check-in patterns, the skewed user check-in frequency and social networks for POI recommendation.”
• 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).”
• DeepMove: Predicting Human Mobility with Attentional Recurrent Networks (2018, methodology): “Early methods for mobility prediction are mostly pattern-based [4, 18, 25, 26, 42].”
• Exploring temporal effects for location recommendation on location-based social networks (2013, background): “Cheng [3] investigated the geographical influence through a multi-center Gaussian model, together with matrix factorization and social influence for location recommendation.”
• Exploiting Geographical Neighborhood Characteristics for Location Recommendation (2014, methodology): “, purchase history [20], click history [5], and check-in history [3].”
• SAR: A sentiment-aspect-region model for user preference analysis in geo-tagged reviews (2015, methodology): “[2] propose latent factor models by incorporating the geographical information using Gaussian distribution.”
• Your neighbors affect your ratings: on geographical neighborhood influence to rating prediction (2014, background): “POI recommendation has attracted significant attention recently with the popularity of LBSN [3, 6, 25, 26].”
• Rank-GeoFM: A Ranking based Geographical Factorization Method for Point of Interest Recommendation (2015, methodology): “For example, geographical coordinates, as an important type of context information which is not available in conventional recommendation tasks, are exploited for POI recommendation [22, 15, 3], since users have a much higher probability to visit nearby POIs [22, 15].”
• Predicting the Next Location: A Recurrent Model with Spatial and Temporal Contexts (2016, methodology): “2011) or the multi-center gaussian model (Cheng et al. 2012).”
• Recommendation (2020, positive): “These models are usually based on matrix factorization for dealing with huge collections of data, and they mostly use check-in data for recommending places [4,9].”