# 04-27 ## Rolling train validate 284 ![](https://i.imgur.com/1caITyL.png) ## Session based/ Sequential models > https://arxiv.org/pdf/1905.01997.pdf > https://github.com/OnYuKang/Recommendation-systems-paperlist > https://yanhuibin315.github.io/2019/07/03/%E5%8A%A8%E6%80%81%E7%9F%A5%E8%AF%86%E5%9B%BE%E8%B0%B1%E8%A1%A5%E5%85%A8%E6%95%B4%E5%90%88/ Sequential -> Session -> Dynamic Graph -> KG with t ![](https://i.imgur.com/7KmQ77a.png) * GRU4Rec * Caser * STAMP: Short-Term Attention/Memory Priority Model for Session-based Recommendation * SASRec * BERT4Rec ![](https://i.imgur.com/eijgePR.png) **GRU -> CNN -> Attention** ## KDD 2019 * Hierarchical Gating Networks for Sequential Recommendation * Streaming Session-based Recommendation * Session-based Social Recommendation via Dynamic Graph Attention Networks * SR-GNN **開始嘗試使用GNN** ## 2020 AAAI * Sequential Recommendation with Relation-Aware Kernelized Self-Attention * Memory Augmented Graph Neural Networks for Sequential Recommendation **都考慮了item共同出現的頻率** ## open dataset * Amazon Beauty * Steam * Movielens * youchoose * yelp