MixGCF: An Improved Training Method for Graph Neural Network-based Recommender Systems.
Tinglin HuangYuxiao DongMing DingZhen YangWenzheng FengXinyu WangJie TangPublished in: KDD (2021)
Keyphrases
- significant improvement
- recommender systems
- high accuracy
- training phase
- segmentation method
- detection method
- computational cost
- probabilistic model
- graph representation
- graph matching
- high precision
- clustering method
- data sets
- dynamic programming
- cost function
- preprocessing
- random walk
- pairwise
- computational complexity
- directed acyclic graph
- similarity measure
- learning algorithm