Toward Unifying Text Segmentation and Long Document Summarization

Benchmark Model Rank Results
text-summarization-on-arxivLodoss-full-large (extractive)#5ROUGE-1: 48.45ROUGE-2: 20.72ROUGE-L: 42.55
text-summarization-on-arxivLodoss-full-base (extractive)#8ROUGE-1: 48.20ROUGE-2: 20.50ROUGE-L: 42.28
text-summarization-on-pubmedLodoss-full-large (extractive)#5ROUGE-1: 49.38ROUGE-2: 23.89ROUGE-L: 44.84
text-summarization-on-pubmedLodoss-full-base (extractive)#7ROUGE-1: 48.93ROUGE-2: 23.51ROUGE-L: 44.40