Open BrambleXu opened 5 years ago
Summary:
KGR的目标是通过corpus里获取的知识来推理missing facts。这个task的一个关键是filter out "irrelevant" facts extracted from corpus。标题里提到的Collaborative Policy Learning指的是通过强化学习的方式,训练两个agents,fact extraction agent从corpora中提取triples,而reasoning agent用于判断生成的fact是否对reasoning有效。
Resource:
Paper information:
Notes:
(额,这个图和caption感觉不搭的样子啊)
These reasoning methods frame the link inference task as a path finding problem over the graph (see Fig. 1 for example)
Model Graph:
为什么 extractor里绿色的original edges也被传入到下面的reasoner了?这个是ground truth吧
Result::
Thoughts:
Next Reading:
Summary:
KGR的目标是通过corpus里获取的知识来推理missing facts。这个task的一个关键是filter out "irrelevant" facts extracted from corpus。标题里提到的Collaborative Policy Learning指的是通过强化学习的方式,训练两个agents,fact extraction agent从corpora中提取triples,而reasoning agent用于判断生成的fact是否对reasoning有效。
Resource:
Paper information:
Notes:
(额,这个图和caption感觉不搭的样子啊)
These reasoning methods frame the link inference task as a path finding problem over the graph (see Fig. 1 for example)
Model Graph:
为什么 extractor里绿色的original edges也被传入到下面的reasoner了?这个是ground truth吧
Result::
Thoughts:
Next Reading: