.png){kind=icon}
返回
Missing link prediction using path and community information
.png){kind=icon}
PRE
AI
DOI：10.1007/s00607-023-01229-y.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Due to the evolving nature of complex networks, link prediction plays a crucial role in exploring likelihood of potential relationships among nodes. There exist a great number of strategies to apply the similarity-based metrics for estimating proximity of nodes in complex networks. In this paper, we propose three new variants - CCPAL3, LPCPA, and GPCPA - for the well-known Common Neighbor and Centrality-based Parameterized Algorithm (CCPA) taking into account 3-hop path, quasi-local path, and global path, respectively. In addition, four novel link prediction strategies based on community detection information, CCPA_CD, CCPAL3_CD, LPCPA_CD and GPCPA_CD, are proposed. Meanwhile, the Jaccard index is extended to three new metrics, i.e., Jaccard_L3, Jaccard_QuasiLoc and Jaccard_Global. Extensive experiments are conducted on thirteen real-world networks. The experimental results indicate that the proposed metrics improve the prediction accuracy measured by AUC and are more competitive on Precision compared to the state-of-the-art link prediction methods.
Keyword:
Link prediction
Local paths
Closeness centrality
Community detection
Complex networks
期刊
C
IF:
2.8
论文数:
2.3K
被引数:
3.5K
机构
引用论文
The Mobile Robot for Garbage Sorting and Handling Based on Machine Vision.png)
2021 IEEE International Conference on Artificial Intelligence, Robotics, and Communication (ICAIRC)
2021 IEEE International Conference on Artificial Intelligence, Robotics, and Communication (ICAIRC)IF0
PREAI
分享收藏Thermoluminescence of glass display from mobile phones for retrospective and accident dosimetry
PREAI
分享收藏分享收藏Reduced -defensin expression is associated with inflammation and not NOD2 mutation status in ileal Crohn's diseaseGut
OAAI
GutIF0
OAAI分享收藏A multi-task neural network for multilingual sentiment classification and language detection on Twitter用于Twitter多语言情感分类和语言检测的多任务神经网络
PREAI
分享收藏
查看更多内容