Hacker News new | ask | show | jobs
by dash2 843 days ago
It's such a shame that recent historical newspapers are closed source in this way. We could learn so much from quantitative analysis of culture if there were downloadable data or at least an API.
4 comments

You can easily find historical newspapers on archive.org. For example:

https://archive.org/details/sim_los-angeles-times_the-los-an...

But automated text analysis needs OCR. There's https://chroniclingamerica.loc.gov but I'm not sure how good its OCR has been.
Well you can, go to your local library and look at microfiche. It's how we used to do it. Primary research. If you're lucky they may have OCR or a tagging system. If not, reels of microfiche
Microfiche is a small flat sheet, microfilm is on reels. "Fiche" being French for card.

The More You Know...

Huh! Til
That's not gonna work at million-scale, though.
I'm old enough to have visited a library and pulled old rolls of microfiche off the shelf to hunt for an archived news article. Being able to pull most of those same records while sitting in my backyard is pretty amazing.

I do fear that too much of the modern era will end up locked behind paywalls and ultimately lost once private equity has no use for it anymore.