Evaluating a Generative Adversarial Framework for Information Retrieval

Abstract

Recent advances in Generative Adversarial Networks (GANs) have resulted in its widespread applications to multiple domains. A recent model, IRGAN, applies this framework to Information Retrieval (IR) and has gained significant attention over the last few years. In this focused work, we critically analyze multiple components of IRGAN, while providing experimental and theoretical evidence of some of its shortcomings. Specifically, we identify issues with the constant baseline term in the policy gradients optimization and show that the generator harms IRGAN's performance. Motivated by our findings, we propose two models influenced by self-contrastive estimation and co-training which outperform IRGAN on two out of the three tasks considered.

References in corpus (11)

• Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
• Dual Learning for Machine Translation
• Adversarial Personalized Ranking for Recommendation
• MaskGAN: Better Text Generation via Filling in the______
• Hindsight Experience Replay
• Attentive Pooling Networks
• The Optimal Reward Baseline for Gradient-Based Reinforcement Learning
• On distinguishability criteria for estimating generative models
• Variance Reduction for Policy Gradient with Action-Dependent Factorized Baselines
• Applying Deep Learning to Answer Selection: A Study and An Open Task
• Notes on Noise Contrastive Estimation and Negative Sampling