Nipers/MDR

An Open-Source repository for multi-domain recommendation.

Related paper list

1. Progressive Layered Extraction (PLE): A Novel Multi-Task Learning (MTL) Model for Personalized Recommendations

   Early stage work, we can see the refinement process of parameters' allocation scheme

2. One Model to Serve All: Star Topology Adaptive Recommender for Multi-Domain CTR Prediction

   Previous SOTA in MDR, simple model designment, easy to read.

3. DisenCDR_Learning_Disentangled_Representations_for_Cross-Domain_Recommendation

   Use Casual graph as analysising tool and inverse preference score as solution to fix domain shift.

4. MAMDR: A Model Agnostic Learning Method for Multi-Domain Recommendation

​ Rejected work, I got the inspiration that origin data can be used as transfer medium

5. Wang et al. - 2022 - CausalInt Causal Inspired Intervention for Multi-Domain Recommendation

   Noah‘ s work this year, our baseline. Use complex casual graph and counterfactual as watercooler moment. Just need to focus on the model design

6. Variational Autoencoders for Collaborative Filtering

   Use VAE as data generator to conduct data-augmentation.

7. Personalized_Transfer_of_User_Preferences_for_Cross-domain_Recommendation

   Propose the theory that we should use aulixary domains with high effiency.

8. DisenCDR_Learning_Disentangled_Representations_for_Cross-Domain_Recommendation

   Propose to disentangle two different kinds of feature.

9. Multi-Sparse-Domain Collaborative Recommendation via Enhanced Comprehensive Aspect Preference Learning

   Purpose two GAN for global and special feature extraction. Inspire me that global features should be indiffieretiable and special features should be opposite.

10. Tenrec: A Large-scale Multipurpose Benchmark Dataset for Recommender Systems

    A dataset built by Tencent.