Siamak at the PAKDD 2024 in Taipei, Taiwan.

28 May 2024

Our external Ph.D. member, Siamak (presenter, pictured) delivered an oral presentation of our accepted paper titled  "Towards Cohesion-Fairness Harmony: Contrastive Regularization in Individual Fair Graph Clustering" by Siamak Ghodsi, Seyed Amjad Seyedi & Eirini Ntoutsi in the research-track of the PAKDD 2024. This paper was a joint work with Siamak’s masters affiliation from the Dept of Computer Engineering, University of Kurdistan, Iran. In his presentation, Siamak discussed the importance of fairness in graph-clustering scenarios. He explained the principles of individual vs. group fairness, their implications, the conflicts with partitioning utility, and how to effectively trade off utility and fairness in unsupervised learning. He then introduced our interpretable contrastive approach to individual fairness and presented our experimental results.

More details on our work can be found under the following link including the paper, codes, implementation, and datasets.