Session-based Recommendation with Graph Neural Networks

By Shu Wu, Yuyuan Tang, Yanqiao Zhu, Liang Wang, Xing Xie, Tien-Ping Tan. AAAI Conference on Artificial Intelligence, 2018

1,914 citations recorded by the source corpus.

The problem of session-based recommendation aims to predict user actions based on anonymous sessions. Previous methods model a session as a sequence and estimate user representations besides item representations to make recommendations. Though achieved promising results, they are insufficient to obtain accurate user vectors in sessions and neglect complex transitions of items. To obtain accurate item embedding and take complex transitions of items into account, we propose a novel method, i.e. Session-based Recommendation with Graph Neural Networks, SR-GNN for brevity. In the proposed method, session sequences are modeled as graphstructured data. Based on the session graph, GNN can capture complex transitions of items, which are difficult to be revealed by previous conventional sequential methods. Each session is then represented as the composition of the global preference and the current interest of that session using an attention network. Extensive experiments conducted on two real datasets show that SR-GNN evidently outperforms the state-of-the-art session-based recommendation methods consistently.

Subfields: Artificial Intelligence, Information Systems

View Paper: https://www.semanticscholar.org/paper/0c7e1338a9c7914a3b9a5bdc42b457b3f272160e

How researchers see this paper

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

In Artificial Intelligence

In Information Systems

Citations in the corpus

A Survey of Graph Neural Networks for Recommender Systems: Challenges, Methods, and Directions (2021, methodology): “In session-based recommendation, most existing works [145, 187, 203, 215, 224] model the session data with a directed graph to capture the item transition pattern.”
Towards Universal Sequence Representation Learning for Recommender Systems (2022, methodology): “Then various kinds of neural network techniques are proposed to encode interaction sequences, such as Recurrent Neural Network (RNN) [13], Transformer [5, 7, 10, 22], Multilayer Perceptron (MLP) [37] and Graph Neural Network (GNN) [2, 27].”
Every Document Owns Its Structure: Inductive Text Classification via Graph Neural Networks (2020, background): “To be speciﬁc, TextGCN (Yao et al., 2019) constructs a single graph with global relations between documents and words, where ﬁne-grained text level word interactions are not considered (Wu et al., 2019; Hu et al., 2019a,b).”
Graph-based knowledge tracing: Modeling student proficiency using graph neural networks (2019, positive): “In recent years, various generalization frameworks and important operations have been developed [3,11,30] with successful results in various research areas, such as social science [13,18] and natural science [2,10,23], and in areas relevant to this work, like time-series recommendation [32].”
Self-Supervised Hypergraph Convolutional Networks for Session-based Recommendation (2020, methodology): “For both datasets, we follow (Wu et al. 2019b; Li et al. 2017) to remove all sessions containing only one item and also remove items appearing less than ﬁve times.”
Graph Neural Networks in Recommender Systems: A Survey (2020, methodology): “Some works [113, 116, 175, 206] calculate the attentive weights between the last item and all the items in the sequence and aggregate the item representations as the global preference, and incorporate it with local preference (i.”
Mining Latent Structures for Multimedia Recommendation (2021, background): “These graph-based recommender systems [14, 41, 43] achieve great success and obtain state-of-the-art performance.”
Multi-behavior Recommendation with Graph Convolutional Networks (2020, methodology): “[39] built graph from session interaction data and adopted GCN to learn representations for session-based recommendation.”
Rethinking the Item Order in Session-based Recommendation with Graph Neural Networks (2019, methodology): “• SR-GNN [38] applies a gated graph convolutional layer [18] to obtain item embeddings, followed by a self-attention of the last item as STAMP does to compute the sequence level”
RecBole: Towards a Unified, Comprehensive and Efficient Framework for Recommendation Algorithms (2020, neutral): “SRGNN [75] AAAI 2019 YOOCHOOSE, DIGINETICA”
Artificial intelligence in recommender systems (2020, background): “GNN is also suited for sequential recommender systems to model the item sequences as a graph [108].”
Sequential Recommender Systems: Challenges, Progress and Prospects (2019, background): “Then, the embeddings of users or items are learned on the graph to embed more complex relations over the whole graph [22].”
XSimGCL: Towards Extremely Simple Graph Contrastive Learning for Recommendation (2022, background): “A great number of recommendation models developed from GNNs have achieved greater than ever performances in different recommendation scenarios [13], [28], [43], [52], [53].”
Contrastive Learning for Sequential Recommendation (2022, methodology): “Since GNN can capture complicated item transition patterns hidden in user sequences, many works attempt to utilize powerful GNN techniques in sequential recommendation tasks, especially when each user sequence can be constructed as a graph.”
STAN: Spatio-Temporal Attention Network for Next Location Recommendation (2021, neutral): “For example, metric embedding algorithms [5, 6], convolutional neural networks [28, 31, 39], reinforcement learning algorithms [24], and graph network [33, 38] are proposed one by one for sequential recommendation.”
A Survey on Graph Neural Networks for Time Series: Forecasting, Classification, Imputation, and Anomaly Detection (2023, background): “Other Applications: The application of GNNs for time series analysis has also been extended to various other ﬁelds, such as ﬁnance [194], [195], on-demand services [41], fraud detection [196], manufacturing [197], and recommender systems [198].”