"Emerald Insight": never getting full text PDFs

terber

When using "Emerald Insight" I never get the full test PDFs.

The translators for other databases (e.g. EBSCO, ScienceDirect, JStor, IEEE XPert) work flawlessly for me.

It also doesn't matter which of my two universities I use.

Example URLs:

http://www.emeraldinsight.com.ezphost.dur.ac.uk/journals.htm?issn=0019-7858&volume=39&issue=2&articleid=1595304&show=html

http://www.emeraldinsight.com.library.myebs.de/journals.htm?issn=0019-7858&volume=39&issue=2&articleid=1595304&show=html

I am using FF 28, Zotero 4.0.19.

Any suggestions?

aurimas

I don't have access to the PDFs on that site, so I can't really fix this. On the page that you linked, is there any link you can right-click -> Save Link As... and download the PDF? If not, how do you get to the PDF? (I'm just seeing purchase/login options)

aurimas

Errr. Nvm, I see the problem. Working on a solution.

(Technical note: unproxied URL is picked up from meta, so the PDF is not downloaded. I'm working on a patch that will proxy all attachment URLs and deproxify metadata URLs. Will submit it soon)

terber

Any progress on this issue? Can I contribute anything to a solution?

Dragomir

My APA format style doesn't work well when downloading form Emerald Insight. The authors' names are not edited properly. Could you please check with that? Thanks!

adamsmith

we'd need more detail than that. Which emerald citation, what's the problem with APA...

DWL-SDCA

Some of the problem is with Emerald's metadata -- even with the metadata the publisher FTPs to databases for indexing purposes. Sometimes the article abstract (while available as text/html on their webpage) is omitted from the metadata. Even within a single article the author name order pattern (both the firstname lastname pattern _and_ the position in the author list) can differ from author to co-author. Sometimes the metadata author names are fewer than that on the printed or PDF version. We never take Emerald metadata at face value but must carefully verify it against the actual published version.

I don't know quite how Zotero does it but I consistently find that the metadata from Zotero translators is more accurate than that we receive directly from the publisher via FTP.

Dragomir

Here is the website link: http://emeraldinsight.com/
When I save in my library any pdf article form this source and then want to cite it in APA it gives me the following pattern:

Gillian Naylor, & Kimberly E. Frank. (2001). The effect of price bundling on consumer perceptions of valuenull. Journal of Services Marketing, 15(4), 270–281. http://doi.org/10.1108/EUM0000000005506

If you pay close attention to the authors' names you can see they are not cited properly, because in APA format style it should be last name then the initial of first name.. examaple:

Naylor G., Frank K. ... and so on

Is there any way you can fix that?
Thanks!

adamsmith

not easily, no. Here's what they give to us:
AU - Gillian Naylor
AU - Kimberly E. Frank

This should be
AU - Naylor, Gillian
AU - Frank, Kimberly E.

You can try reporting that to Emerald (this is their RIS export).
You can manually fix it after import by clicking on the white rectangle after the authors' names.

Dragomir

Thanks, that works! =)

DWL-SDCA

Note that the word "null" is appended to the title of Emerald journal articles.

adamsmith

that's pretty much universally the case? I suppose we can try to clean it up. It's in the RIS, so not much we can do otherwise.

DWL-SDCA

I believe that after September 2014 all current and backfile metadata titles have had that string appended to the titles. We modified our parser that translates MODS xml to the SafetyLit format so that "null" is stripped from the final word of the title of articles from journals having an Emerald DOI prefix. I have emailed and regular mailed the Emerald A&I office about this at least 3 times. They replied saying that they are looking into it. That was back in February.

adamsmith

thanks, that's very useful. We can strip that, too, and restrict that operation the the emeraldinsight domain so we're not catching any poor article actually ending with "null".

DWL-SDCA

The null problem is that "null" is appended to the last word of the title. So, if the last word of the article's title was "Null" the final word in the metadata would be Nullnull. (There isn't a space before null.) This is becoming "Who's on first?"

adamsmith

yeah, but we use the same translator for a wide range of sites, so we can't assume all data has that problem.

aurimas

not easily, no. Here's what they give to us:
AU - Gillian Naylor
AU - Kimberly E. Frank

This should be
AU - Naylor, Gillian
AU - Frank, Kimberly E.

We should treat all names as personal names in RIS when it is called from other translators (which should mean that it's called from web translator). This is a frequent problem and it's rarely an institutional name. We should still do it correctly when importing RIS files directly.

DWL-SDCA

The problem is that Emerald will often provide

AU - Naylor, Gillian
AU - Kimberly E. Frank
AU - Smith, John
AU - George James

The next time I encounter this I will report the DOI here.

adamsmith

that wouldn't be a problem -- in similar cases (e.g. when scraping metadata from a webpage header) we just check for the presence of a comma and use that to decide in which format the name is.
As aurimas indicates, the only issue is that this will get institutional names wrong -- and I think he's correct that those are quite rare in the types of databases we'd use RIS from. I'm not 100% convinced by what he suggests, but it would definitely increase the share of correctly imported data. (My objection is that it would essentially "punish" places with good data -- by getting institutional authors wrong -- in order to accommodate those with crappy data.)

adamsmith

Emerald now strips the "null" at the end of titles.