Matches in SemOpenAlex for { <https://semopenalex.org/work/W3217502908> ?p ?o ?g. }
- W3217502908 abstract "Recently, Graph Neural Networks (GNNs) have proven their effectiveness for recommender systems. Existing studies have applied GNNs to capture collaborative relations in the data. However, in real-world scenarios, the relations in a recommendation graph can be of various kinds. For example, two movies may be associated either by the same genre or by the same director/actor. If we use a single graph to elaborate all these relations, the graph can be too complex to process. To address this issue, we bring the idea of pre-training to process the complex graph step by step. Based on the idea of divide-and-conquer, we separate the large graph into three sub-graphs: user graph, item graph, and user-item interaction graph. Then the user and item embeddings are pre-trained from user and item graphs, respectively. To conduct pre-training, we construct the multi-relational user graph and item graph, respectively, based on their attributes. In this paper, we propose a novel Reinforced Attentive Multi-relational Graph Neural Network (RAM-GNN) to the pre-train user and item embeddings on the user and item graph prior to the recommendation step. Specifically, we design a relation-level attention layer to learn the importance of different relations. Next, a Reinforced Neighbor Sampler (RNS) is applied to search the optimal filtering threshold for sampling top-k similar neighbors in the graph, which avoids the over-smoothing issue. We initialize the recommendation model with the pre-trained user/item embeddings. Finally, an aggregation-based GNN model is utilized to learn from the collaborative relations in the user-item interaction graph and provide recommendations. Our experiments demonstrate that RAM-GNN outperforms other state-of-the-art graph-based recommendation models and multi-relational graph neural networks." @default.
- W3217502908 created "2021-12-06" @default.
- W3217502908 creator A5002591086 @default.
- W3217502908 creator A5022309774 @default.
- W3217502908 creator A5036357902 @default.
- W3217502908 creator A5055097463 @default.
- W3217502908 creator A5069093595 @default.
- W3217502908 creator A5079118288 @default.
- W3217502908 creator A5079669827 @default.
- W3217502908 date "2021-11-27" @default.
- W3217502908 modified "2023-09-26" @default.
- W3217502908 title "Pre-training Recommender Systems via Reinforced Attentive Multi-relational Graph Neural Network" @default.
- W3217502908 cites W1516061453 @default.
- W3217502908 cites W1533230146 @default.
- W3217502908 cites W2017710739 @default.
- W3217502908 cites W2039613841 @default.
- W3217502908 cites W2108920354 @default.
- W3217502908 cites W2127795553 @default.
- W3217502908 cites W2153579005 @default.
- W3217502908 cites W2184957013 @default.
- W3217502908 cites W2219888463 @default.
- W3217502908 cites W2295739661 @default.
- W3217502908 cites W2509893387 @default.
- W3217502908 cites W2604314403 @default.
- W3217502908 cites W2624407581 @default.
- W3217502908 cites W2766453196 @default.
- W3217502908 cites W2884134047 @default.
- W3217502908 cites W2945623882 @default.
- W3217502908 cites W2949687195 @default.
- W3217502908 cites W2949972983 @default.
- W3217502908 cites W2950393809 @default.
- W3217502908 cites W2950975304 @default.
- W3217502908 cites W2962767366 @default.
- W3217502908 cites W2962992837 @default.
- W3217502908 cites W2963323306 @default.
- W3217502908 cites W2963341956 @default.
- W3217502908 cites W2963403868 @default.
- W3217502908 cites W2963869731 @default.
- W3217502908 cites W2964015378 @default.
- W3217502908 cites W2964051675 @default.
- W3217502908 cites W2964311892 @default.
- W3217502908 cites W2964341035 @default.
- W3217502908 cites W2966779056 @default.
- W3217502908 cites W2970597249 @default.
- W3217502908 cites W2986711944 @default.
- W3217502908 cites W2995448904 @default.
- W3217502908 cites W3023045989 @default.
- W3217502908 cites W3031029974 @default.
- W3217502908 cites W3035399171 @default.
- W3217502908 cites W3043239945 @default.
- W3217502908 cites W3045200674 @default.
- W3217502908 cites W3080510905 @default.
- W3217502908 cites W3098087397 @default.
- W3217502908 cites W3099386565 @default.
- W3217502908 cites W3100278010 @default.
- W3217502908 cites W3106439716 @default.
- W3217502908 cites W3106445281 @default.
- W3217502908 cites W3114654929 @default.
- W3217502908 cites W3138089704 @default.
- W3217502908 cites W3139159537 @default.
- W3217502908 cites W3153911428 @default.
- W3217502908 cites W3156051320 @default.
- W3217502908 cites W3192966765 @default.
- W3217502908 cites W3193441787 @default.
- W3217502908 cites W3195672100 @default.
- W3217502908 cites W3205731400 @default.
- W3217502908 cites W2900252255 @default.
- W3217502908 doi "https://doi.org/10.48550/arxiv.2111.14036" @default.
- W3217502908 hasPublicationYear "2021" @default.
- W3217502908 type Work @default.
- W3217502908 sameAs 3217502908 @default.
- W3217502908 citedByCount "0" @default.
- W3217502908 crossrefType "posted-content" @default.
- W3217502908 hasAuthorship W3217502908A5002591086 @default.
- W3217502908 hasAuthorship W3217502908A5022309774 @default.
- W3217502908 hasAuthorship W3217502908A5036357902 @default.
- W3217502908 hasAuthorship W3217502908A5055097463 @default.
- W3217502908 hasAuthorship W3217502908A5069093595 @default.
- W3217502908 hasAuthorship W3217502908A5079118288 @default.
- W3217502908 hasAuthorship W3217502908A5079669827 @default.
- W3217502908 hasBestOaLocation W32175029081 @default.
- W3217502908 hasConcept C119857082 @default.
- W3217502908 hasConcept C132525143 @default.
- W3217502908 hasConcept C154945302 @default.
- W3217502908 hasConcept C21569690 @default.
- W3217502908 hasConcept C23123220 @default.
- W3217502908 hasConcept C41008148 @default.
- W3217502908 hasConcept C557471498 @default.
- W3217502908 hasConcept C80444323 @default.
- W3217502908 hasConceptScore W3217502908C119857082 @default.
- W3217502908 hasConceptScore W3217502908C132525143 @default.
- W3217502908 hasConceptScore W3217502908C154945302 @default.
- W3217502908 hasConceptScore W3217502908C21569690 @default.
- W3217502908 hasConceptScore W3217502908C23123220 @default.
- W3217502908 hasConceptScore W3217502908C41008148 @default.
- W3217502908 hasConceptScore W3217502908C557471498 @default.
- W3217502908 hasConceptScore W3217502908C80444323 @default.
- W3217502908 hasLocation W32175029081 @default.
- W3217502908 hasLocation W32175029082 @default.
- W3217502908 hasOpenAccess W3217502908 @default.