Datasets:
Tasks:
Text Generation
Modalities:
Text
Formats:
parquet
Languages:
English
Size:
10M - 100M
Tags:
ocr
License:
Wondering About Coverage and Ad Removal
#4
by
AshleyLL
- opened
I’ve recently come across the dell-research-harvard/newswire dataset on Huggingface, which appears to be based on the same source material as PleIAs/US-PD-Newspapers.
I would appreciate it if you could kindly help clarify a few points:
1.How does the PleIAs/US-PD-Newspapers dataset compare to dell-research-harvard/newswire in terms of data quality and coverage?
2.What approach can be used to remove advertisements from the newspaper pages in PleIAs/US-PD-Newspapers ?
3.When comparing PleIAs/US-PD-Newspapers with datasets like dell-research-harvard/newswir for the same years, which one covers a wider range of newspaper sources?
Thanks!