linkvec

This repository provides a reference implementation of linkvec This repository contains following files:

1. ./emb/embedding.emb : Contains the embedding for the given graph
2. ./graph/fb-food-pages.edgelist : Contains the edgelist for facebook food page dataset
3. ./src/main.py : Contains driver program for embedding generation
4. ./src/linkprediciton.py : Contains driver program for link predicition problem
5. ./src/algorithm.py : Contains methods for edge embedding.
6. ./src/linkvectorizer.py : Contains implementation of link vectorizer object.

Embedding Generation

Basic Usage

Example

To run linkvec on random network, execute the following command from the project home directory:
python src/main.py --input graph/fb-food-pages.edgelist --output emb/random.emb

Options

You can check out the other options available to use with linkvec using:
python src/main.py --help

Input

The supported input format is an edgelist:

node1_id_int node2_id_int <weight_float, optional>

The graph is assumed to be undirected and unweighted by default. These options can be changed by setting the appropriate flags.

Output

The intermediatery output file has n+1 lines for a graph with n vertices. The first line has the following format:

num_of_nodes dim_of_representation

The next n lines are as follows:

node_id dim1 dim2 ... dimd

the final edge embedding would be <dim1 ,dim2, ... dimd>

Link Prediction

To predict links one can use src/linkprediction.py

The given implementation is for computing the accuracy. Rest of the metrics can be computed accordingly.

Citation

@article{TRIPATHI2021,
title = {Network embedding based link prediction in dynamic networks},
journal = {Future Generation Computer Systems},
year = {2021},
issn = {0167-739X},
doi = {https://doi.org/10.1016/j.future.2021.09.024},
url = {https://www.sciencedirect.com/science/article/pii/S0167739X21003708},
author = {Shashi Prakash Tripathi and Rahul Kumar Yadav and Abhay Kumar Rai},
keywords = {Network embedding, Link prediction, Similarity measures, Network features, Feature learning, Biased random walk, Dynamic networks},
}