dqwang122 / HeterSumGraph

Code for ACL2020 paper "Heterogeneous Graph Neural Networks for Extractive Document Summarization"
244 stars 52 forks source link

No 'summary' field in train.lable.jsonl #10

Closed nlpmc closed 4 years ago

nlpmc commented 4 years ago

Hi,

I encounter the same problem (#2 ) that there is no summary field in the training file. The code of this project is really clean and nice. Although one can solve this by using the provided features file, I want to conduct another experiment based on this code repo.

So could you please provide the complete training file that will also help the guys who want to prepare the graph feature by themselves? Thank you so much!

dqwang122 commented 4 years ago

In fact, we do not need 'summary' key during the training phase, which has been claimed in README. Since this code is only for extractive summarization, we just predict 0/1 for each sentence in the document. If you want to perform the code in another dataset, you can extract labels (the true supervised signal) by the code in #7

nlpmc commented 4 years ago

It's very kind of you to provide the code for extracting labels. Thanks!