This is an implementation of the MVE model for multi-view network embedding (An Attention-based Collaboration Framework for Multi-View Network Representation Learning).
Our codes rely on two external packages, which are the Eigen package and the GSL package.
The Eigen package is used for matrix operations. To run our codes, users need to download the Eigen package and modify the package path in the makefile.
The GSL package is used to generate random numbers. After installing the package, users also need to modify the package path in the makefile.
After installing the two packages and modifying the package paths, users may go to the "mve" folder and use the makefile to compile the codes.
The MVE model receives a multi-view network and a set of labeled nodes as input.
Each view of the multi-view network is described by a single file. The files of different views should have the same prefix, and the indices should start from 0 to K-1, where K is the number of views. For example, users may describe a multi-view network with three files, "view_0", "view_1", "view_2", where the prefix is "view_" and the number of views is 3. Each view file contains several lines, with each line representing an edge in that view. The format of each line is: "u v w", meaning that there is an edge from node "u" to node "v" and the weight is "w".
The labeled nodes are listed in another file. This file contains several lines, where each line gives the labels of a node. The format of a line is: "node_name label_name_1 label_name_2", which starts with the name of the node, followed by the names of different labels the node has.
A toy dataset is provided in the "data/toy/" folder. For the DBLP, Flickr and Youtube datasets, they are available here.
To run the MVE model, users may directly use the example script (run.sh) we provide.
If you have any questions about the codes, please feel free to contact us.
Meng Qu, [email protected]
@inproceedings{qu2017attention,
title={An Attention-based Collaboration Framework for Multi-View Network Representation Learning},
author={Qu, Meng and Tang, Jian and Shang, Jingbo and Ren, Xiang and Zhang, Ming and Han, Jiawei},
booktitle={Proceedings of the 2017 ACM on Conference on Information and Knowledge Management},
pages={1767--1776},
year={2017},
organization={ACM}
}