Metapath-guided Heterogeneous Graph Neural Network for Intent Recommendation 个人总结
写在前面:为方便阅读,尽量使用中文总结,对于翻译无把握或专有词汇,在中文后附上原文字段.
另,CSDN对markdown 导入、latex支持还有进步空间啊,最完整的体验劳请移步GitHub拉下来再用Typora打开?(无耻引流开始了):my github
0. 原作信息
@inproceedings{fan2019metapath, title={Metapath-guided Heterogeneous Graph Neural Network for Intent Recommendation}, author={Fan, Shaohua and Zhu, Junxiong and Han, Xiaotian and Shi, Chuan and Hu, Linmei and Ma, Biyu and Li, Yongliang}, booktitle={Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery \& Data Mining}, pages={2478--2486}, year={2019} }
1. Contributions 本文贡献
-
提出基于词嵌入的预处理(propose a uniform term embedding mechanism);
-
为意图推荐任务,设计基于异构图上metapath的模型MEIRec( design a metapath-guided heterogeneous Graph Neural Network to learn the embeddings of objects);
-
线下、线上实验都证明了模型的有效性。
2. Backgrounds 背景信息
2.1 意图推荐 (e-commerce intent recommendation)
例子:
-
输入数据类型:
-
特征数据 attribute data
-
交互数据 interaction data
-
-
与传统推荐系统的区别:
-
加入交互数据(点击等
-