RA-EGES
1.运用场景
大规模商品推荐。
2.创新点
Based on years of practical experience in Taobao,we design an effective heuristic method to construct the item graph from the behavior history of one billion users in Taobao;
Propose three embedding methods,BGE,GES,and EGES, to learn embeddings of two billion items in Taobao。
3.算法原理
3.1 网络框架
![](https://img2020.cnblogs.com/blog/1722723/202005/1722723-20200505110900051-815248850.png)
3.2 EGES
![](https://img2020.cnblogs.com/blog/1722723/202005/1722723-20200505111014902-857864066.png)
4.算法理解
使用图模型和side information来生成item向量。
人生忽如茶,莫辜负茶、汤和好天气。