Personalized Ranking Metric Embedding for Next New POI Recommendation

By Shanshan Feng, Xutao Li, Yi-feng Zeng, Gao Cong, Y. Chee, Quan Yuan. International Joint Conference on Artificial Intelligence, 2015

539 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/ac3090517977ce4189dde194f8c3f9ea9367a66c

How researchers see this paper

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

1. Xiangnan He
2. Chen Gao
3. Shoujin Wang
4. Xiang Wang
5. Yongfeng Zhang
6. Jiacheng Li
7. Xu Chen
8. Hongzhi Yin

Citations in the corpus

• A Survey on Session-based Recommender Systems (2019, background): “…and low-order sequential dependencies Usually ignore long-term and higher-order dependencies, the rigid order assumption is too strong [13],[23],[28],[56],[95],[139], [154] GenerativeprobabilisticSBRSs Ordered sessions with high-order dependencies Good at modelling high-order and collective…”
• Where to Go Next: Modeling Long- and Short-Term User Preferences for Point-of-Interest Recommendation (2020, methodology): “…al. (2013) apply a localized region constraint by factorizing personalized Markov Chain (FPMC) (Rendle, Freudenthaler, and Schmidt-Thieme 2010), and Feng et al. (2015) consider the personalized sequential information as well as model the geographical influence by calculating the distance between…”
• Sequential Recommender Systems: Challenges, Progress and Prospects (2019, background): “The former one directly calculates the transition probability based on the explicit observations [3], while the latter first embeds the Markov chains into an Euclidean space and then calculates the transition probabilities between interactions based on their Euclidean distance [2].”
• Where to Go Next: A Spatio-Temporal Gated Network for Next POI Recommendation (2019, methodology): “Feng et al.(2015) proposed a personalized ranking metric embedding method (PRME) to model personalized check-in sequences for next POI recommendation.”
• A Survey of Graph Neural Networks for Recommender Systems: Challenges, Methods, and Directions (2021, background): “This motivates the problem of next-POI or successive POI recommendation [46, 102, 189, 213].”
• Self-Attentive Sequential Recommendation (2018, methodology): “PRME [47], HRM [43], Fossil [21]) have been outperformed on similar datasets by baselines among those above, we omit comparison against them.”
• Sequence-Aware Recommender Systems (2018, methodology): “ 1, Article 1. Publication date: February 2018. Sequence-Aware Recommender Systems 1:19 in which the evolution of the student knowledge is formalized by a Hidden Markov Model. The goal of Feng et al. [30] is personalized POI recommendation based on the last-N points of interest visited by the user. ”
• Predicting the Next Location: A Recurrent Model with Spatial and Temporal Contexts (2016, methodology): “And PFMC improves the performance greatly comparing with TF. PFMC-LR and PRME achieve further improvement with via incorporating distance information.”
• Sequential Recommender System based on Hierarchical Attention Networks (2018, methodology): “[Chen et al. , 2012] and [Feng et al. , 2015] used metric embedding to project items into points in a low-dimension Euclidean space for play list prediction and successive location recommendation.”
• Learning Graph-based POI Embedding for Location-based Recommendation (2016, background): “PRME proposed by Feng et al. [6] is also the typical one which exploits pair-wise ranking scheme.”
• Sequential Recommendation with User Memory Networks (2018, unknown): “Beyond e-commerce, sequential recommendation has also been applied to various application scenarios such as POI recommendation [3, 4], music recommendation [1, 8, 29], browsing recommendation [35], etc.”
• Evaluation of session-based recommendation algorithms (2018, methodology): “d method from [27], which we also considered in the experiments. Another family of sequence modeling approaches relies on distributeditemrepresentations, e.g., in the form of latent Markov embeddings [4, 5, 13, 58] or distributional embeddings [2, 10, 15, 43, 51, 56, 64]. Embeddings are dense, lower”
• Rank-GeoFM: A Ranking based Geographical Factorization Method for Point of Interest Recommendation (2015, unknown): “10, 8, 26], and currently-visited POI [4, 6].”
• Translation-based Recommendation (2017, methodology): “Recently, Personalized Ranking Metric Embedding (PRME) [Feng et al. , 2015] was proposed to improve FPMC by learning two metric spaces: one for measuring user-item affinity and another for sequential continuity.”
• STAN: Spatio-Temporal Attention Network for Next Location Recommendation (2021, background): “Meanwhile, such a system can also provide users with the convenience to decide where to go and how to plan the day, based on previous visits as well as current status [5, 7, 9, 23, 41].”
• Time Interval Aware Self-Attention for Sequential Recommendation (2020, methodology): “PRME [6], HRM [27], Fossil [8], GRU4Rec[12]) have been outperformed by the baselines above, so we omit comparison against them.”