Bias Reduction via Cooperative Bargaining in Synthetic Graph Dataset Generation

TL;DR

This paper introduces a cooperative bargaining method to generate synthetic graph datasets with balanced representation across various metrics, aiming to reduce selection bias and improve benchmarking reliability for graph neural networks and processing frameworks.

Contribution

The work presents a novel cooperative bargaining approach for creating unbiased synthetic graph datasets with even metric distribution, enhancing dataset quality for benchmarking.

Findings

Synthetic datasets with balanced metrics improve GNN benchmarking accuracy.

The method effectively reduces selection bias in graph datasets.

Enhanced datasets lead to more reliable evaluation of graph processing techniques.

Abstract

In general, to draw robust conclusions from a dataset, all the analyzed population must be represented on said dataset. Having a dataset that does not fulfill this condition normally leads to selection bias. Additionally, graphs have been used to model a wide variety of problems. Although synthetic graphs can be used to augment available real graph datasets to overcome selection bias, the generation of unbiased synthetic datasets is complex with current tools. In this work, we propose a method to find a synthetic graph dataset that has an even representation of graphs with different metrics. The resulting dataset can then be used, among others, for benchmarking graph processing techniques as the accuracy of different Graph Neural Network (GNN) models or the speedups obtained by different graph processing acceleration frameworks.