ItemRank: A Random-Walk Based Scoring Algorithm for Recommender Engines

By Marco Gori, Augusto Pucci. International Joint Conference on Artificial Intelligence, 2007

353 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/78a00de20967358b5d4dc53a0d6ab1cf59418a68

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. Xiang Wang
3. Chen Gao
4. Fuli Feng
5. Tat-seng Chua
6. Liqiang Nie
7. Zhiyong Cheng
8. Shiwen Wu

Citations in the corpus

• Graph Neural Networks in Recommender Systems: A Survey (2020, methodology): “In fact, using collaborative signals to improve representation learning in recommender systems is not a new idea that originated from GNN [40, 68, 75, 183, 202].”
• GraphSAIL: Graph Structure Aware Incremental Learning for Recommender Systems (2020, methodology): “performance of “batch mode” in a sliding window manner. Specifically, days[T,T +13]of data are used for training, dayT +14 of data is used for validation, and dayT +15 of data is for testing, whereT ∈[1,6]. On the other hand, the model trained by different incremental learning strategies starts with”
• PageRank beyond the Web (2014, background): “Then the ItemRank scores [Gori and Pucci, 2007] are the solutions of: where D R are column sums of the rating matrix.”
• Interest-aware Message-Passing GCN for Recommendation (2021, background): “Early approaches infer indirect preference by random walks in the graph to provide recommendation [7, 10, 11].”
• TriRank: Review-aware Explainable Recommendation by Modeling Aspects (2015, methodology): “TagRW enhances ItemRank [6] by incorporating tags into building the item–item graph and performing an additional random walk on user–user graph.”
• Multi-behavior Recommendation with Graph Convolutional Networks (2020, methodology): “Some works [1, 8, 12, 25] apply random-walk based methods to graph-based recommendation.”
• A Comprehensive Survey of Neighborhood-based Recommendation Methods (2011, methodology): “A recommendation approach, based on the PageRank algorithm for ranking Web pages [11], is ItemRank [26].”
• Community detection in graphs (2009, methodology): “This method accounts for the fact that a point may belong to two or more clusters at the same time and is widely used in pattern recognition.”
• Neural Graph Collaborative Filtering (2019, methodology): “s that meet the desired properties. 3.2 Graph-Based CF Methods Another line of research [12, 24, 39] exploits the user-item interaction graph to infer user preference. Early efforts, such as ItemRank [7] and BiRank [12], adopt the idea of label propagation to capture the CF effect. To score items fo”
• Ieee Transactions on Knowledge and Data Engineering, Manuscropt Id 1 Social Recommendation with Cross-domain Transferable Knowledge (unknown): “ItemRank [40], a random walk-based scoring algorithm, was used to rank products according to expected user preferences.”
• Improving Graph Collaborative Filtering with Neighborhood-enriched Contrastive Learning (2022, background): “Early studies [1, 6] extract the structure information through random walks in the graph.”
• Recommender Systems (2016, unknown): “Such a graph is also referred to as the correlation graph [232].”
• LightGCN: Simplifying and Powering Graph Convolution Network for Recommendation (2020, methodology): “at improves the embedding learning. 5.2 Graph Methods for Recommendation Another relevant research line is exploiting the user-item graph structure for recommendation. Prior efforts, such as ItemRank [11] and BiRank [16], use the label propagation mechanism to directly propagate user preference scor”
• Collaborative Filtering beyond the User-Item Matrix (2014, methodology): “Examples include the use of walks to improve item-item similarity matrix for item-based CF [Gori and Pucci 2007; Yildirim and Krishnamoorthy 2008] to infer social trust relationship for trust-based recommendation [Jamali and Ester 2009a, 2009b].”
• Collaborative filtering recommender systems (2013, background): “…ratings, the MSD similarity compared to PC similarity has one shortcoming that it does not capture negative correlations between user preferences or the appreciation of different items but having such negative correlations can improve the rating prediction accuracy (Gori and Pucci, 2007).”