Contents

🚗 1. fundamental part of RS

• notes on shallow model
  • notes on AMF
  • notes on SVD++
  • notes on FISM
• notes on deep model
  • notes on FM
  • notes on NFM
  • notes on Wide&Deep
  • notes on Deep Crossing
  • notes on NCF

🚗 2. adversarial training (adversarial perturbations)

• paper notes on APR
• paper notes on SACRA

🚗 3. adversarial learning (generative adversarial network)

• paper notes on APOIR
• paper notes on Cascading DQN
• paper notes on DASO
• paper notes on ESRF
• paper notes on Geo-ALM
• paper notes on LARA
• paper notes on MFGAN

• Title:Adversarial Personalized Ranking for Recommendation[🌟first train MF with BPR, and then further optimize it under APR framework]

• Title: Adversarial Learning to Compare: Self-Attentive Prospective Customer Recommendation in Location based Social Networks[🌟prospective customer recommendation, recommend by making comparisons among users’ historical check-ins with adversarial training]

• Title: Adversarial Point-of-Interest Recommendation[🌟geographical and social influence are also incorporated]

• Title: Generative Adversarial User Model for Reinforcement Learning Based Recommendation System[🌟Using the estimated user behavior model φ and the corresponding reward function r as the simulation environment, then use reinforcement learning to obtain a recommendation policy.]

• Title: Deep Adversarial Social Recommendation[🌟bidirectional mappings between the user-item interactions (item domain) and user-user connections (social domain)]

• Title: Enhancing Social Recommendation with Adversarial Graph Convolutional Networks[🌟alternative neighborhood generation, neighborhood denoising, and attention-aware social recommendation, GCN based]

• Title: Geo-ALM: POI Recommendation by Fusing Geographical Information and Adversarial Learning Mechanism[🌟fuse geographical features with generative adversarial networks(GAN)]

• Title: LARA: Attribute-to-feature Adversarial Learning for New-item Recommendation[🌟aim to optimize a matching problem between existing users and a virtual user profile generated from attributes of the targeting new-item(item cold-start problem)]

• Title: Sequential Recommendation with Self-AttentiveMulti-Adversarial Network[🌟a Transformer-based generator taking user behavior sequences as input to recommend the possible next items, multiple factor-specific discriminators to evaluate the generated sub-sequence from the perspectives of different factors]