IRGAN: A Minimax Game for Unifying Generative and Discriminative Information Retrieval Models

By Jun Wang, Lantao Yu, Weinan Zhang, Yu Gong, Yinghui Xu, Benyou Wang, P. Zhang, Dell Zhang. 2017

686 citations recorded by the source corpus.

This paper provides a unified account of two schools of thinking in information retrieval modelling: the generative retrieval focusing on predicting relevant documents given a query, and the discriminative retrieval focusing on predicting relevancy given a query-document pair. We propose a game theoretical minimax game to iteratively optimise both models. On one hand, the discriminative model, aiming to mine signals from labelled and unlabelled data, provides guidance to train the generative model towards fitting the underlying relevance distribution over documents given the query. On the other hand, the generative model, acting as an attacker to the current discriminative model, generates difficult examples for the discriminative model in an adversarial way by minimising its discrimination objective. With the competition between these two models, we show that the unified framework takes advantage of both schools of thinking: (i) the generative model learns to fit the relevance distribution over documents via the signals from the discriminative model, and (ii) the discriminative model is able to exploit the unlabelled data selected by the generative model to achieve a better estimation for document ranking. Our experimental results have demonstrated significant performance gains as much as 23.96% on Precision@5 and 15.50% on MAP over strong baselines in a variety of applications including web search, item recommendation, and question answering.

Subfields: Molecular Biology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

Citations in the corpus

Synthesizing Tabular Data using Generative Adversarial Networks (2018, background): “Other GAN applications include information retrieval [45], dialogue systems [27], and speech processing [32].”
KDGAN: Knowledge Distillation with Generative Adversarial Networks (2018, methodology): “Similar to other two-player frameworks [49], the naive adaptation requires a large number of training instances and epochs [15], which is difficult to satisfy in practice [4].”
Learning to Collaborate for Question Answering and Asking (2018, background): “Besides, our work differs from (Wang et al., 2017) in that our question generation model is a generative model while theirs is actually a discriminative matching model.”
GraphGAN: Graph Representation Learning with Generative Adversarial Nets (2017, methodology): “In addition, several prior works also try to preserve spe-ciﬁc properties during the learning process, such as global structures (Wang, Cui, and Zhu 2016), community structures (Wang et al. 2017c), group information (Chen, Zhang, and Huang 2016), and asymmetric transitivity (Ou et al. 2016).”
A Review on Generative Adversarial Networks: Algorithms, Theory, and Applications (2020, background): “Field Subfield Method Super-resolution SRGAN [63], ESRGAN [64], Cycle-in-Cycle GANs [65], SRDGAN [66], TGAN [67] DR-GAN [68], TP-GAN [69], PG2 [70], PSGAN [71], Image synthesis and manipulation APDrawingGAN [72], IGAN [73], Image processing and computer vision introspective adversarial networks [74]”
CoaCor: Code Annotation for Code Retrieval with Reinforcement Learning (2019, background): “Various kinds of machine-machine collaboration mechanisms have been studied in many scenarios [12, 15, 28, 55, 59].”
Adversarial Distillation for Efficient Recommendation with External Knowledge (2018, background): “[44] applied the idea of generative adversarial network into the filed of information retrieval (IR), and demonstrate its effectiveness in three specific tasks including: document retrieval, recommendation system and question answering.”
Recommendation system based on deep learning methods: a systematic review and new directions (2019, methodology): “This method was later extended by Lian et al. (2017), and Wang et al. (2017) to the cross-domain recommendation.”
Adversarial and Contrastive Variational Autoencoder for Sequential Recommendation (2021, methodology): “GAN-based models such as IRGAN [25], CFGAN [2] can be used for recommendation.”
A Survey on Generative Adversarial Networks: Variants, Applications, and Training (2020, background): “) [157] for question-answer selection, Resume-GAN [158] for talent-job fit, Natural Language GAN (NLG) [159] and Fake-GAN [160] for review detection and generation, Information Retrieval GAN (IR-GAN) [161] and Personalized Search GAN (PS-GAN) [162] for information retrieval, Adversarial Reward Learn”
PACRR: A Position-Aware Neural IR Model for Relevance Matching (2017, background): “Learning to rank models, like the recent IRGAN model (Wang et al., 2017), rely on handcrafted features to encode query document interactions, e.g., the relevance scores from unsupervised ranking models.”
Bias and Debias in Recommender System: A Survey and Future Directions (2020, background): “An example is the adaptive negative sampler [40], [41], [42], [43], which aims to over-sample the “difficult” instances in order to increase learning speed, even though the resultant loss function will differ significantly from the original.”
KBGAN: Adversarial Learning for Knowledge Graph Embeddings (2017, background): “IRGAN (Wang et al., 2017) is a recent work which combines two categories of information retrieval models into a discrete GAN framework.”
Artificial intelligence in recommender systems (2020, background): “…that learns the distribution over documents and try to generate relevant documents that look like the ground truth to fool the discriminative model, and a discriminative model that aims to classify the ground-truth documents from the generated ones as an opponent to the generative model [102].”
EXS: Explainable Search Using Local Model Agnostic Interpretability (2018, positive): “…networks have been widely successful in an array of learning tasks including vision, speech and NLP. Recently, the IR community has also seen successful application of deep learning models in document representation [3, 9], query modeling [4, 19, 20] and design of ranking models [6, 10, 11, 16].”