Thanks for the responses; keep them coming, if you have other ideas.
It's hard to demarcate domains, but my userbase consists largely of
librarians and liberal arts faculty.
I wasn't at all aware of Microsoft Academic Search. Their content looks
thorough, though it doesn' include books:
But based on the few searches I tried, it looks to be horrible at de-duping.
I like the idea of using Zotero or Mendeley but that would put the
burden of managing the data on me.
My hope was either to find some reasonable data ready-made (MS Academic
Search shows promise) or to relegate that to my users by making the
process useful in other ways (ORCID).
Dan, can you use ORCID IDs with Zotero? (Though that might create
another hoop to jump through.)
Another tack in my case might be to aim for an indicative, rather an
exhaustive, list of works. That would require more human brain than API,
but would decrease the size of the data.
On 05/21/2015 12:23 AM, Dan Scott wrote:
> Good timing for this discussion!
> On Wed, 20 May 2015 at 17:03 Laura Robbins <[log in to unmask]> wrote:
>> If you're looking to compile your own data, Zotero is a great way to
>> do it and provides an API. I recently moved our faculty publications
>> database into it.
> We're embarking on a self-compilation exercise and are going the Zotero
> route as well, relying on the power of student labour to take the
> bibliography section of each faculty member's CV and add them to Zotero
> group libraries.
> I'm cobbling together a super-simple Flask-on-PostgreSQL web app that will
> take the Zotero export as RIS, migrate it into a relational database, and
> enable us to clean up / enrich the data from there (
> https://github.com/dbs/ris2web if you want to peek at it's very early, very
> much in progress state).
> One thing I've learned is that I realllllllly wish is that Zotero offered
> linked data authority control, so that I could assert that this "Smith,
> Jane" is either different or the same as that "Smith, Jane" at the time of
> creating or editing the bibliography entry, but for now we're going to do
> the deduping & URI addition after the fact. The URIs will be exposed as
> RDFa in the HTML of course.
>> If you're looking to compile data, though I hate to suggest it, is
>> there an API for google scholar? It's not a perfect resource, but for
>> humanities and social sciences, I've found some publications for our
>> faculty that they hadn't given me themselves.
> There is no API for Google Scholar, alas. There is an API for Microsoft
> Academic Search at http://academic.research.microsoft.com/About/Help.htm#4
> which might be of interest. Haven't tried it myself.
> The Mendeley API looks quite interesting as well. Wasn't aware of that,
> even--thanks Bee!