magiasn / commonsenseerl_emnlp_2019 Goto Github PK

您好，请问代码里面的ntn模型在哪儿可以找到，可以提供下吗？
(model/nyt/ntn/NeuralTensorNetwork_2007.pt)

你好！想请教一下，具体是怎么替换SGNN中的event embeding呀

Hi,
Thanks for the nice work.
I wonder if there is any code for Script generation using your event representation?
In the paper, the representations were evaluated on script event prediction. However I wondered if we can generate script?

你好，关于NTY语料的预处理我有几个问题：
1、 我现在已经有了LDC2008T19从1987-2007的原始语料，但里面有许多非‘故事’的文本，比如company report，请问这些文本需要过滤么
2、 代码中只对原始文本做了分段处理，在运行ollie之前文本还需要做tokenize等进一步nlp处理么
3、 预处理代码好像并没有做共指解析，存在许多代词如“he”，“it”等该如何处理

As mentioned in knowitall/ollie#9, the Open IE tool internally used MaltParser for parsing, which can not handle unicode correclty.

So the extraction script in https://github.com/MagiaSN/CommonsenseERL_EMNLP_2019/blob/master/preproc/OpenExtract.scala should add the following code to handle coding exception.

import scala.io.Codec
import java.nio.charset.CodingErrorAction

implicit val codec = Codec("UTF-8")
codec.onMalformedInput(CodingErrorAction.REPLACE)
codec.onUnmappableCharacter(CodingErrorAction.REPLACE)