Adversarial Graph Embeddings for Fair Influence Maximization over Social Networks
Authors: Moein Khajehnejad, Ahmad Asgharian Rezaei, Mahmoudreza Babaei, Jessica Hoffmann, Mahdi Jalili, Adrian Weller
IJCAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | While there are typically trade-offs between fairness and influence maximization objectives, our experiments on synthetic and real-world datasets show that our approach dramatically reduces disparity while remaining competitive with state-of-the-art influence maximization methods. |
| Researcher Affiliation | Academia | Moein Khajehnejad1 , Ahmad Asgharian Rezaei2 , Mahmoudreza Babaei3 , Jessica Hoffmann4 , Mahdi Jalili2 and Adrian Weller5,6 1Monash University 2RMIT University 3Max Planck Institute for Software Systems 4The University of Texas at Austin 5University of Cambridge 6Alan Turing Institute |
| Pseudocode | Yes | Algorithm 1 ADVERSARIAL GRAPH EMBEDDING: Adversarially trains the Embedder function E with a Discriminator function D. ... Algorithm 2 FAIR SELECTION: Selects |S| initial seeds attempting to choose the most influential ones. |
| Open Source Code | Yes | For code and details of parameter tunings please refer to: https://github.com/Ahmadreza401/fair influen max |
| Open Datasets | Yes | The real dataset is the Rice University Facebook dataset, collected by [Mislove et al., 2010], which represents the friendship relations between students of Rice University. |
| Dataset Splits | No | The paper describes using synthetic and real-world datasets and mentions splitting into groups (e.g., Group A and Group B for sensitive attributes) but does not provide explicit training, validation, and test splits (e.g., percentages or counts). |
| Hardware Specification | No | No specific hardware details (e.g., GPU/CPU models, memory) used for running the experiments were provided in the paper. |
| Software Dependencies | No | The paper does not provide specific software dependencies with version numbers. |
| Experiment Setup | No | The paper mentions parameters like 'activation probability' and 'connection probabilities' for synthetic data, and details of how communities are defined for the Rice Facebook dataset. However, it does not explicitly provide hyperparameter values for the Adversarial Graph Embeddings model itself (e.g., learning rate, batch size, number of epochs for the autoencoder or discriminator). |