Quote from gkelliott on May 7, 2019, 11:54 am

I see several requests on Zotero forums for a translator to be built that would pull proper metadata off NewsBank and other related sites. Just doesn't exist yet. I have a ton of these articles I downloaded in the past and always anticipated having to key the citations. Was just hoping for a snappy import of the future finds. 

 

Quote from Donna Cox Baker on May 9, 2019, 5:59 pm

Hi, gkelliott. I have spoken with Zotero about what it can do for sites like FamilySearch, which picks up a webpage for every entry, as you saw for some other sites. It appears that the first thing that has to happen is for a data provider to open its metadata to view. Zotero describes that here:

https://www.zotero.org/support/dev/exposing_metadata

And the data necessary to create citations has to be there and has to be set up with all the needed pieces accessible as discrete units. Newspapers.com does not appear to have cited things at the article level--at least not on the newspapers I've looked at. They offer you a page with fully searchable text. And they will let you clip the piece of the page you want. But they've left it to you to tease out the discrete elements--the title of an article, etc. 

I was fortunate in graduate school to be using databases that had created all of the citation information at the article level--databases like "America's Historical Newspapers" and "Nineteenth-Century American Newspapers." (I might not be remembering the titles exactly--but something like those.) It did the work for me there. It worked for most of the databases our university used, and I think it is because the scholarly community has long been aware of Zotero and other tools that can scrape citation data.

The providers of genealogy data might need to be made aware of Zotero and what they can do to make their information "scrapable." I sent the link above to the CIO of FamilySearch about a month ago, to see what they could do. He was very interested and gracious about it. He passed it to his team for review. I hope it will turn into an improvement there. 

I encourage all of us to pass the link (https://www.zotero.org/support/dev/exposing_metadata) to the companies developing the genealogy tools we are using. Ask them to make their tools Zotero-ready. In doing so, I think they will also be making them ready for a number of other tools that grab citations. I know that EndNote does, and I think there is a Google product that scrapes citation information. All require an exposed metadata--and a metadata that separates out all of the citation elements into discrete parts, rather than giving us just a block of text that is a suggested citation (though we will take that gratefully, if it is all that is offered).

It is also possible that some sources can give up their citation metadata, if handled just the right way--as we figured out with gmails recently. Take a look at that post.

I think, once the genealogy community is fully aware of Zotero and what it can do, our database companies will be thinking about how to jump on this bandwagon. So I hope all of us will be asking them to.

If any of the rest of you have had success with the other databases, please weigh in.

And, when you grab the web citation from Newspapers.com, change the data type to newspaper. Then at least a few pieces of information have been typed for you, which can select and drag to the proper fields. You migth find it easier, though, to just type from scratch. 

Or, here's another thought. Find another source that has the articles cited and grab from there. I know that won't always be possible--or even maybe most of the time. Try ChroniclingAmerica. It won't always come out perfectly, but it does a lot of the work for you.

Thanks so much for raising this. I wish I had an easy answer....

 

Quote from gkelliott on May 12, 2019, 11:34 pm

Thanks Donna.  With those suggestions, I'll get a flow that works.