OpenANE: The first Open source framework specialized in Attributed Network Embedding (ANE)

We reproduce several ANE (Attributed Network Embedding) as well as PNE (Pure Network Embedding) methods in one unified framework, where they all share the same I/O and downstream tasks. We start this project based on OpenNE that mainly integrates PNE methods under one unified framework.
OpenANE not only integrates those PNE methods that consider pure structural information, but also provides the state-of-the-art ANE methods that consider both structural and attribute information during embedding.

Authors: Chengbin HOU chengbin.hou10@foxmail.com & Zeyu DONG 2018

Motivation

In many real-world scenarios, a network often comes with node attributes such as paper metadata in a citation network, user profiles in a social network, and even node degrees in any pure networks. Unfortunately, PNE methods cannot make use of attribute information that may further improve the quality of node embeddings.
From engineering perspective, by offering more APIs to handle attribute information in graph.py and utils.py, OpenANE shall be very easy to use for embedding an attributed network. Except attributed networks, OpenANE can also deal with pure networks by calling PNE methods, or by assigning node degrees as node attributes and then calling ANE methods. Therefore, to some extent, ANE methods can be regarded as a generalization of PNE methods.

Methods

ABRW, SAGE-GCN, SAGE-Mean, ASNE, TADW, AANE, DeepWalk, Node2Vec, LINE, GraRep, AttrPure, AttrComb
Note: all NE methods in this framework are unsupervised, and so does not require any label during embedding phase.

For more details of each method, please have a look at our paper https://arxiv.org/abs/1811.11728
And if you find ABRW or this framework is useful for your research, please consider citing it.

Usages

Requirements

pip install -r requirements.txt

Python 3.6.6 or above is required due to the new print(f' ') feature

To obtain node embeddings as well as evaluate the quality

python src/main.py --method abrw --emb-file emb/cora_abrw_emb --save-emb --task lp_and_nc

To have an intuitive feeling in node embeddings

python src/vis.py --emb-file emb/cora_abrw_emb --label-file data/cora/cora_label.txt

Testing (Cora)

Parameter Settings

In this testing, we use the default parameters

AANE_lamb	AANE_maxiter	AANE_rho	ABRW_alpha	ABRW_topk	ASNE_lamb	AttrComb_mode	GraRep_kstep	LINE_negative_ratio	LINE_order	Node2Vec_p	Node2Vec_q	TADW_lamb	TADW_maxiter	batch_size	dim	dropout	epochs	label_reserved	learning_rate	link_remove	number_walks	walk_length	weight_decay	window_size	workers
0.05	10	5	0.8	30	1	concat	4	5	3	0.5	0.5	0.2	10	128	128	0.5	100	0.7	0.001	0.1	10	80	0.0001	10	24

Testing Results

Link Prediction and Node Classification tasks:

method	AUC	Micro-F1	Macro-F1	Time
aane	0.8158	0.7263	0.6904	26.48
abrw	0.9290	0.8721	0.8603	48.94
asne	0.7842	0.6076	0.5649	69.67
attrcomb	0.9111	0.8444	0.8284	60.32
attrpure	0.7857	0.7349	0.7039	0.49
deepwalk	0.8499	0.8100	0.8021	75.15
grarep	0.8936	0.7669	0.7607	10.31
line	0.6945	0.5873	0.5645	259.02
node2vec	0.7938	0.7977	0.7858	29.42
sagegcn	0.8929	0.7780	0.7622	207.49
sagemean	0.8882	0.8057	0.7902	183.65
tadw	0.9005	0.8383	0.8255	10.73

2D Visualization task

Steps: Cora -> NE method -> node embeddings -> PCA -> 2D vis
The different colors indicate different ground truth labels.

Other Datasets

More well-prepared (attributed) network datasets are available at NetEmb-Datasets

Your Own Dataset

*--------------- Structural Info (each row) --------------------*
adjlist: node_id1 node_id2 node_id3 ... (neighbors of node_id1)
or edgelist: node_id1 node_id2 weight (weight is optional)
*--------------- Attribute Info (each row) ---------------------*
node_id1 attr1 attr2 ...
*--------------- Label Info (each row) -------------------------*
node_id1 label1 label2 ...

Parameter Tuning

For different dataset, one may need to search the optimal parameters instead of taking the default parameters. For the meaning and suggestion of each parameter, please see main.py.

Contribution

We highly welcome and appreciate your contribution on fixing bugs, reproducing new ANE methods, etc. And we hope this OpenANE framework would become influential on both academic research and industrial usage.

References

todo...