realcatking / aane_python

Accelerated Attributed Network Embedding, SDM 2017

AANE_Python

• Accelerated Attributed Network Embedding, SDM 2017
• A General Embedding Framework for Heterogeneous Information Learning in Large-Scale Networks, TKDD, 2018

Installation

• Requirements

1. numpy
2. scipy

• Usage

1. cd AANE_Python
2. pip install -r requirements.txt
3. python Runme.py

Input and Output

• Input: dataset such as "BlogCatalog.mat" and "Flickr.mat"
• Output: Embedding.mat, with "H_AANE" denotes the attributed network embedding, and "H_Net" denotes the network embedding

Code in Python

from AANE_fun import AANE_fun
H = AANE_fun(Net,Attri,d)
H = AANE_fun(Net,Attri,d,lambd,rho)
H = AANE_fun(Net,Attri,d,lambd,rho,'Att')
H = AANE_fun(Net,Attri,d,lambd,rho,'Att',splitnum)

• H is the joint embedding representation of Net and Attri;
• Net is the weighted adjacency matrix;
• Attri is the node attribute information matrix with row denotes nodes;
• splitnum is the number of pieces we split the SA for limited cache.
• Python 3.6.3 or 2.7.13 is recommended.

Reference in BibTeX:

@conference{Huang-etal17Accelerated,
Author = {Xiao Huang and Jundong Li and Xia Hu},
Booktitle = {SIAM International Conference on Data Mining},
Pages = {633--641},
Title = {Accelerated Attributed Network Embedding},
Year = {2017}}

@article{Huang-etal18A,
Title = {A General Embedding Framework for Heterogeneous Information Learning in Large-Scale Networks},
Author = {Xiao Huang and Jundong Li and Na Zou and Xia Hu},
Booktitle = {ACM Transactions on Knowledge Discovery from Data},
Volume = {12},
Year = {2018}}

Code for Distributed Computing

from AANE_fun_distri import AANE_fun
H = AANE_fun(Net,Attri,d)
H = AANE_fun(Net,Attri,d,lambd,rho)
H = AANE_fun(Net,Attri,d,lambd,rho,'Att')
H = AANE_fun(Net,Attri,d,lambd,rho,'Att',splitnum, worknum)

• H is the joint embedding representation of Net and Attri;
• Net is the weighted adjacency matrix;
• Attri is the node attribute information matrix with row denotes nodes;
• splitnum is the number of pieces we split the SA for limited cache;
• worknum is the number of worker.

The function for distributed computing could only be run on macOS with Python 3.6.3 recommended.