Alapfogalmak
Proposing RCoCo for collective link prediction across multiplex networks in Riemannian space.
Kivonat
研究提出了RCoCo模型,用于在黎曼空间中进行多重网络的集体链接预测。该模型利用曲率感知图注意力网络(κ-GAT)协同地处理网络内和网络间行为,以解决几个挑战:合作的内部和网络间行为、表示空间选择、稀缺锚定用户学习等。通过广泛实验验证了RCoCo的有效性。
Statisztikák
Facebook: 422,291 nodes, 3,710,789 links
TwitterA: 669,198 nodes, 12,749,257 links
TwitterB: 5,167 nodes, 164,660 links
Foursquare: 5,240 nodes, 76,972 links
DBpediaCH: 66,469 nodes, 153,929 links
DBpediaEN: 98,125 nodes, 237.674 links
AMiner: 26,386 nodes, 273,476 links
DBLP: 24.352 nodes.316.565 links
表1显示了Facebook、TwitterA、TwitterB、Foursquare、DBpediaCH、DBpediaEN、AMiner和DBLP数据集的几何特征。
Idézetek
"Link prediction typically studies the probability of future interconnection among nodes with the observation in a single social network."
"In RCoCo...we propose to study a challenging yet practical problem of Geometry-aware Collective Link Prediction across Multiplex Network."
"We argue that intra-link prediction boosts inter-link prediction and vice versa."