Codes for link prediction in multi-relational networks (.m files)

• Xi Wang (xiwang@eecs.ucf.edu)
• Gita Sukthankar (gitars@eecs.ucf.edu)

This repository contains the codes (written in Matlab format) of 8 unsupervised link predictors (and their corresponding weighted version) used in LPSF model (Link Prediction using Social Features) and 2 diffusion-based link prediction models: LPDP (Link Prediction using Diffusion Pro- cess) and LPDM (Link Prediction using Diffusion Maps).

---

8 unsupervised link predictors include:

• Preferential Attachment (PA);
• Adamic/Adar (AA);
• Common Neighbors (CN);
• Resource Allocation Index (RA)
• Jaccard's Coefficient (JC),
• PropFlow
• Inverse Path Distance
• PageRank

---

Please cite our corresponding papers if you use this material in any form in your publication:

@inproceedings{Sukthankar-Xi-ASONAM2013,
	author = {Wang, Xi and Sukthankar, Gita},
	title = {Link Prediction in Multi-relational Collaboration Networks},
  	booktitle = {Proceedings of the IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining},
  	address = {Niagara Falls, Canada},
  	month = {Aug},
  	year = {2013},
  	pages = {1445--1447}
}

OR the following accepted paper.

@incollections{Sukthankar-Xi-LNSN2014,
	author = {Wang, Xi and Sukthankar, Gita},
	title = {Link Prediction in Heterogeneous Collaboration Networks},
  	booktitle = {Lecture Notes in Social Networks},
  	editor = {Alhajj, Reda, Glässer, Uwe}
  	publisher  = {Springer},
  	year = {2014}
}