← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 6486.i · 生成模型:GPT-6

图网络连接聚合与更新

理解等级:中级

图神经网络利用节点间关系聚合邻域信息,再更新节点表示。用于推荐时,用户、物品或知识实体的关系可能提供直接交互之外的信息;效果仍取决于图的质量、关系含义和传播范围,图中相关不等于因果。

这张卡片看不懂?

已有反馈 · 0 条

    还没有人提交反馈。

    在整书笔记中的位置

    原书出处 · 1 处

    • · 查看原书依据

      Graph Neural Network (GNN) is an emerging approach that facilitates object feature extraction from the knowledge graphs and provides adequate information to the RS. GNN consists of an aggregator and an updater. Aggregators are classified into a relation-aware subgraph, relation-unaware, and relation-aware attentive types. Updates are categorized as single-interaction, multi-interaction, and context-only [45].

    参考来源

    - 位置:《Artificial Intelligence and Data Science in Recommendation System: Current Trends, Technologies, and Applications》第03章《深度推荐与评估指标》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Artificial Intelligence and Data Science in Recommendation System: Current Trends, Technologies, and Applications/_epub-src/OEBPS/Text/9789815136746-C3.html) - 结论依据:摘录中关于图网络连接聚合与更新的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。 - 原始内容:Graph Neural Network (GNN) is an emerging approach that facilitates object feature extraction from the knowledge graphs and provides adequate information to the RS. GNN consists of an aggregator and an updater. Aggregators are classified into a relation-aware subgraph, relation-unaware, and relation-aware attentive types. Updates are categorized as single-interaction, multi-interaction, and context-only [45].