I'm Hussain Hussain, a third year PhD student at TU Graz (Institute of Interactive Systems and Data Science) and Know-Center, supervised by Dr. Roman Kern.

I'm mainly interested in graph mining and currently I focus on the impact of network structural properties on the output of graph learning models. See my academic CV or the publication list below.

Feel free to reach out

• hussain@tugraz.at
• Twitter: sqr_hussain
• Github: sqrhussain

Publications

[Google Scholar page]

• Adversarial Inter-Group Link Injection Degrades the Fairness of Graph Neural Networks
  Hussain Hussain, Meng Cao, Sandipan Sikdar, Denis Helic, Elisabeth Lex, Markus Strohmaier, Roman Kern.
  To appear at IEEE ICDM 2022 as a short paper.
  [preprint] [code]
• Effective Use of BERT in Graph Embeddings for Sparse Knowledge Graph Completion.
  Xinglan Liu, Hussain Hussain, Houssam Razouk, Roman Kern.
  ACM SAC'22, 790-793 (2022).
  https://dl.acm.org/doi/10.1145/3477314.3507031
• The interplay between communities and homophily in semi-supervised classification using graph neural networks.
  H Hussain, T Duricic, E Lex, D Helic, R Kern.
  Applied Network Science 6, 80 (2021).
  https://doi.org/10.1007/s41109-021-00423-1
  [preprint] [code]
• Structack: Structure-based Adversarial Attacks on Graph Neural Networks.
  H Hussain, T Duricic, E Lex, D Helic, M Strohmaier, R Kern.
  Proceedings of the 32st ACM Conference on Hypertext and Social Media (2021).
  [preprint] [code]
• Should We Embed in Chemistry? A Comparison of Unsupervised Transfer Learning with PCA, UMAP, and VAE on Molecular Fingerprints
  Mario Lovrić, Tomislav Đuričić, Han TN Tran, Hussain Hussain, Emanuel Lacić, Morten A Rasmussen, Roman Kern
  Pharmaceuticals 14 (8), 758 (2021).
• On the Impact of Communities on Semi-supervised Classification Using Graph Neural Networks.
  H Hussain, T Duricic, E Lex, R Kern, D Helic.
  International Conference on Complex Networks and Their Applications, 15-26 (2020).
  [preprint] [code]
• Empirical Comparison of Graph Embeddings for Trust-Based Collaborative Filtering.
  T Duricic, H Hussain, E Lacic, D Kowald, D Helic, E Lex
  International Symposium on Methodologies for Intelligent Systems, 181-191 (2020).
  [preprint]

TU Graz Data Team

Each semester, we participate in data challenges together in the TU Graz data team.

You can participate in a challenge within the data team as a student project in KDDM2 (706.715) VU course in the winter semster or NLP (706.230) VU course in the summer semster.

Join us now for some data science, drinks and pizza ;)

News Timeline

Good News

• 22-09-01 - Our paper titled "Adversarial Inter-Group Link Injection Degrades the Fairness of Graph Neural Networks" got accepted at IEEE ICDM 2022.
• 21-12-16 - Our paper titled "Effective Use of BERT in Graph Embeddings for Sparse Knowledge Graph Completion" was accepted for publication in ACM SAC '22 conference.
• 21-09-07 - Our paper titled "The Interplay between Communities and Homophily in Semi-supervised Classification Using Graph Neural Networks." was accepted at Applied Network Science special issue. This is an extension of the CompNets2020 paper. [preprint] [code]
• 21-07-09 - Our paper "Structack: Structure-based Adversarial Attacks on Graph Neural Networks" got accepted at ACM Hypertext 2021. [preprint] [code]
• 20-09-30 - Our paper "On the Impact of Communities on Semi-supervised Classification Using Graph Neural Networks" got accepted at Complex Networks 2020. [preprint] [code]

Bad News

• 22-05-19 - Our paper on fairness attacks on GNNs got rejected at KDD 2022.
• 21-09-07 - I created this webpage.
• 21-05-16 - Our paper "Structack: Structure-based Adversarial Attacks on Graph Neural Networks" was rejected at ACM SIGKDD 2021.
• 20-06-05 - Our paper "On the Impact of Communities and Hubs on Semi-supervised Classification Using Graph Neural Networks" got rejected at ECML-PKDD 2020.
• 20-02-28 - Our abstract "Semi-supervised Classification on Directed Graphs" was rejected at ELLIS Workshop on Geometric and Relational Deep Learning 2020.
• 19-10-01 - Our paper "Semi-Supervised Classification for Directed Graphs with Bi-Directional Graph Convolutional Networks" was rejected at NeurIPS 2019 Graph Representation Learning workshop.