Wen Xiao

Hi, I have a quick follow-up question on this point. For 'oracle entities', which NER tool did you used for extacting oracle entities from the reference summary? Thanks a lot!!

It is one of the input of model.predict() for all the models in models.py, with parameter name 'length_limit'

I c, you mean the length of the input documents. I simply created sub-datasets based on the length of the input documents, and evaluated them separately.

For each dataset (pubmed and arxiv), I split the ENTIRE test set into FOUR sub-datasets based on **the length of the input documents **(e.g. 0-2000, 2000-4000,4000-6000,>6000 for pubmed). I can...

Hi Janice, Sorry that I just saw the issue, we simply use sections to represent each topic, i.e. each section itself is a topic segment.

Hi there, Running the training/testing process does not require java, it is only needed in one evaluation metric, 'METEOR', if you don't want to set up the environments for java,...

Hi, It was caused by a minor mistake when adapting the rouge_papier. You could solve it by changing 'rouge_papier' to 'rouge_papier_v2' in line 16&18 in the file rouge_papier_v2/rouge_papier_v2/wrapper.py, and then...