The source code for the paper GraphConfRec: A Graph Neural Network-Based Conference Recommender System

This directory contains the code necessary to run GraphConfRec. We provide the implementation of various recommendation models, along with evaluation scripts. The repository is organised as follows:

• data/: contains the necessary raw dataset files for GraphConfRec and stores the processed data, including trained models
• notebooks/: contains Jupyter Notebooks for exporing the SciGraph, WikiCfP, and H5 Index datasets
• src/: contains the implementation of the recommendation models, evaluation and data preprocessing scripts

1. Download the raw data files (see Datasets).
2. Preprocess the raw data files and crawl external datasets (see Data preprocessing).
3. Preprocess training data, train and evaluate recommendation models (see Models).

We implemented the following recommendation models:

The code was tested running under Python 3.6.8, with the following packages installed (along with their dependencies):

• numpy==1.16.2
• pandas==0.24.1
• scipy==1.2.1
• networkx==2.2
• beautifulsoup4==4.7.1
• tensorflow==2.0.0
• tensorflow-gpu==2.0.0
• torch==1.3.0
• torch-cluster==1.4.5
• torch-geometric==1.3.2
• torch-scatter==1.4.0
• torch-sparse==0.4.3
• transformers==2.1.1
• Flask==1.1.1
• Flask-SQLAlchemy==2.4.1
• SQLAlchemy==1.3.1

In addition, CUDA 10.0 was used.

The MIT license is applied to the provided source code. For the datasets, please check the licensing information:

• See https://scigraph.springernature.com/explorer/license/ for SciGraph.
• See http://www.wikicfp.com/cfp/data.jsp for WikiCfP.

Parts of the code were originally forked from:

• GraphSAGE
• Advancing GraphSAGE with A Data-driven Node Sampling
• AS-GCN in Tensorflow
• Graph Attention Networks
• Heterogeneous Graph Neural Network

We owe many thanks to the authors of the different models for making their codes available.

If you use this code in your research, please cite the paper as follows:

@misc{iana2021graphconfrec,
      title={GraphConfRec: A Graph Neural Network-Based Conference Recommender System}, 
      author={Andreea Iana and Heiko Paulheim},
      year={2021},
      eprint={2106.12340},
      archivePrefix={arXiv},
      primaryClass={cs.IR}
}