论文网址:Multi-Channel Graph Neural Network for Entity Alignment (aclanthology.org)
论文代码:https:// github.com/thunlp/MuGNN
英文是纯手打的!论文原文的summarizing and paraphrasing。可能会出现难以避免的拼写错误和语法错误,若有发现欢迎评论指正!文章偏向于笔记,谨慎食用
目录
2.3. Preliminaries and Framework
2.4.1. Rule Inference and Transfer
2.5. Multi-Channel Graph Neural Network
2.5.2. Multi-Channel GNN Encoder
2.6.3. Impact of Two Channels and Rule Transfer
2.6.4. Impact of Seed Alignments
1. 心得
(1)是比较容易理解的论文
2. 论文逐段精读
2.1. Abstract
①Limitations of entity alignment: structural heterogeneity and limited seed alignments
②They proposed Multi-channel Graph Neural Network model (MuGNN)
2.2. Introduction
①Knowledge graph (KG) stores information by directed graph, where the nodes are entity and the edges denote relationship
②Mother tongue information usually stores more information:
(作者觉得KG1的Jilin会对齐KG2的Jilin City,因为他们有相似的方言和连接的长春。这个感觉不是一定吧?取决于具体模型?感觉还是挺有差别的啊这俩东西,结构上也没有很相似)
③To solve the problem, it is necessary to fill in missing entities and eliminate unnecessary ones
2.3. Preliminaries and Framework
2.3.1. Preliminaries
(1)KG
①Defining a directed graph , which contains entity set , relation set and triplets
②Triplet
(2)Rule knowledge
①For rule , , it means there are
(3)Rule Grounding
①通过上面的递推,实体可以找到更进一步的关系
(4)Entity alignment
①Alignments in two entities:
②Alignment relation:
2.3.2. Framework
①Workflow of MuGNN:
(1)KG completion
①Adopt rule mining system AMIE+
(2)Multi-channel Graph Neural Network
①Encoding KG in different channels
2.4. KG Completion
2.4.1. Rule Inference and Transfer
2.4.2. Rule Grounding
①比如从KG2中找到关系,就可以补充到KG1中去
2.5. Multi-Channel Graph Neural Network
2.5.1. Relation Weighting
①They will generate a weighted relationship matrix
②They construct self attention adjacency matrix and cross-KG attention adjacency matrix for each channel
(1)KG Self-Attention(这个是为了补齐)
①Normalized connection weights:
where contains self loop and denotes the neighbors of
② denotes the attention coefficient between two entities:
where and are trainable parameters
(2)Cross-KG Attention(这个是为了修剪,是另一个邻接矩阵)
①Pruning operation :
if is true then it will be 1 otherwise 0, denotes inner product similarity measure
2.5.2. Multi-Channel GNN Encoder
①Propagation of GNN:
and they chose as ReLU
②Multi GNN encoder:
where denotes the number of channels
③Updating function:
④Pooling strategy: mean pooling
2.5.3. Align Model
①Embedding two KG to the same vector space and measure the distance to judge the equivalence relation:
where , , and are negative pairs in the original sets, and are margin hyper-parameters separating positive and negative entity and relation alignments
②Triplet loss:
③ denotes the true value function for triplet :
then it can be recursively transformed into:
where is the embedding size
④The overall loss:
2.6. Experiment
2.6.1. Experiment Settings
(1)Datasets
①Datasets: DBP15K (contains DBPZH-EN(Chinese to English), DBPJA-EN (Japanese to English), and DBPFREN (French to English)) and DWY100K (contains DWY-WD (DBpedia to Wikidata) and DWY-YG (DBpedia to YAGO3))
②Statistics of datasets:
③Statistics of KG in datasets:
(2)Baselines
①MTransE
②JAPE
③GCN-Align
④AlignEA
(3)Training Details
①Training ratio: 30% for training and 70% for testing
②All the embedding size: 128
③All the GNN layers: 2
④Optimizer: Adagrad
⑤Hyperparameter:
⑥Grid search to learning rate in {0.1,0.01,0.001}, L2 in {0.01,0.001,0.0001}, dropout rate in {0.1,0.2,0.5}. They finally got 0.001,0.01,0.2 optimal each
2.6.2. Overall Performance
2.6.3. Impact of Two Channels and Rule Transfer
①Module ablation:
2.6.4. Impact of Seed Alignments
①Ratio of seeds:
2.6.5. Qualitative Analysis
①Two examples of how the rule works:
2.7. Related Work
Introduces some related works
2.8. Conclusions
They aim to further research word ambiguity
3. 知识补充
3.1. Adagrad Optimizer
(1)补充学习:Deep Learning 最优化方法之AdaGrad - 知乎 (zhihu.com)
4. Reference
Cao, Y. et al. (2019) 'Multi-Channel Graph Neural Network for Entity Alignment', Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, doi: 10.18653/v1/P19-1140