This dataset contains various processed text features extracted from articles and web retrieval systems. It includes human-generated content, web-retrieved content, and entity extraction results using different Named Entity Recognition (NER) models.
And the Web Ret 4 method is the best to lower ECD score:
Use abstract -> Query Google
Use Human Story Fetched -> Ranked the sentences
abstract: The main abstract of… See the full description on the dataset page:
https://huggingface.co/datasets/Rajarshi-Roy-research/web_ret_exp_100.
