知识嵌入增强的对比推荐模型
|
|
谢涛,葛慧丽,陈宁,汪晓锋,李延松,黄晓峰
|
Knowledge embedding-enhanced contrastive recommendation model
|
|
Tao XIE,Huili GE,Ning CHEN,Xiaofeng WANG,Yansong LI,Xiaofeng HUANG
|
|
| 表 3 基于NDCG@K指标的推荐算法性能比较 |
| Tab.3 Performance comparison of recommendation algorithms based on NDCG@K metrics |
|
| 模型 | NDCG@10 | | NDCG@20 | | NDCG@30 | | Yelp2018 | Amazon-Book | MIND | | Yelp2018 | Amazon-Book | MIND | | Yelp2018 | Amazon-Book | MIND | | KGAT[26] | 0.035 7 | 0.061 5 | 0.031 7 | | 0.043 2 | 0.073 9 | 0.044 2 | | 0.051 8 | 0.081 5 | 0.051 6 | | KGIN[27] | 0.036 8 | 0.062 4 | 0.039 5 | | 0.046 2 | 0.074 8 | 0.052 7 | | 0.053 4 | 0.082 9 | 0.059 1 | | CKAN[28] | 0.036 4 | 0.060 8 | 0.036 8 | | 0.044 1 | 0.072 0 | 0.049 9 | | 0.052 6 | 0.081 2 | 0.056 5 | | KGCL[32] | 0.038 5 | 0.063 2 | 0.042 5 | | 0.049 3 | 0.079 3 | 0.055 1 | | 0.057 3 | 0.087 4 | 0.062 0 | | MGDCF[29] | 0.035 1 | 0.068 5 | 0.042 2 | | 0.051 6 | 0.083 1 | 0.056 4 | | 0.061 8 | 0.091 2 | 0.062 4 | | 本研究方法 | 0.044 1 | 0.074 8 | 0.055 1 | | 0.056 3 | 0.091 9 | 0.070 1 | | 0.065 0 | 0.100 6 | 0.077 3 |
|
|
|