Multi-Interest Network with Dynamic Routing for Recommendation at Tmall
主要思想是用胶囊网络做多兴趣提取
胶囊网络为什么适用于多兴趣提取呢
（1）首先多兴趣网络会通过初次的向量求和，相当于来说去在某些方向上累加兴趣倾向
（2）然后通过加大与和向量相似程度高的增加权重进而来提取重要兴趣，不断迭代去寻找累加兴趣，基本可以让最大的兴趣有最好的表达
（3）为什么可以多兴趣的较好地表达呢
从实验结果来看，多兴趣可以较好地表达，为啥呢???
1、研究目标
to represent one user with multiple vectors encoding the different aspects of the user’s interests.
We propose the Multi-Interest Network with Dynamic routing (MIND) for dealing with user’s diverse interests in the matching stage.
2、研究方法
Specifically, we design a multi-interest extractor layer based on capsule routing mechanism, which is applicable for clustering historical behaviors and extracting diverse interests.
Furthermore, we develop a technique named label-aware attention to help learn a user representation with multiple vectors.
3、论文架构
参考文献：
1、如何刻画用户的多样兴趣——MIND network阅读笔记
https://zhuanlan.zhihu.com/p/68897114
2、论文原文
https://arxiv.org/pdf/1904.08030.pdf
3、Multi-Interest Network with Dynamic Routing for Recommendation at Tmall-MIND多兴趣动态路由推荐
https://blog.csdn.net/itbigpig/article/details/103876984
4、推荐系统遇上深度学习(七十四)-[天猫]MIND：多兴趣向量召回
