Skip to content

Releases: Huffon/factsumm

v0.1.5: resolve SentencePiece dependency issue

01 Jan 16:46
Compare
Choose a tag to compare
  • resolve SentencePiece dependency issue

v0.1.4: Remove Flair, OpenIE modules & Improve Triple logic

01 Jan 16:01
Compare
Choose a tag to compare
  • remove Flair, we only use huggingface for named entity recognition
  • remove OpenIE-based open-scheme triple score (too unstable)
  • apply hf's official named entity recognition aggregation strategy
  • fix BERTScore length issues by applying iteration with summary lines

v0.1.2: Update with BERTScore

13 May 16:56
Compare
Choose a tag to compare

Update BERTScore based Module (See Sec 4.1 from https://arxiv.org/pdf/2005.03754.pdf)

>>> factsumm = FactSumm()
>>> factsumm.calculate_bert_score(article, summary)
BERTScore Score
Precision: 0.9151781797409058
Recall: 0.9141832590103149
F1: 0.9150083661079407

v0.1.1: Initial pypi release

12 May 15:42
Compare
Choose a tag to compare

Currently FactSumm supports the following methods:

  • NER + RE based Triple Module
  • QG + QA based Module
  • OpenIE based Triple Module
  • ROUGE based Module