CLAW IR Minimum Viable Product

islandora-interest-groups / Islandora-IR-Interest-Group

An IG focused on improving Islandora as an IR platform

13 stars 17 forks source link

CLAW IR Minimum Viable Product #31

Open bryjbrown opened 8 years ago

bryjbrown commented 8 years ago

What features should CLAW have in order to be considered a usable IR? List your required features below.

bryjbrown commented 8 years ago

IP and chronological embargoes for objects and datastreams
Support for PDF-based objects
Support for binary-style objects (object behaves independently of file format)
Support for compound objects

manez commented 8 years ago

I'd definitely like to hear opinions about the binary-style objects handling. We don't technically have that in the 7.x stack (although Binary Object SP is in Labs), which has always seem like a gap to me. A nice opportunity to have CLAW offer a concrete improvement, perhaps.

BCDigIR commented 8 years ago

Support for both self-deposit and batch ingest pathways
Support for basic SEO (ex. metatags, XML sitemap)
Method for gathering usage statistics/metrics on individual objects/datastreams

dmoses commented 8 years ago

I'll have to look at the CLAW MVP, but I think the minimum would be:

support for descriptive metadata
support for an associated binary stream -- it could be a PDF or some other type? -- the system could use the mimetype of the binary to provide a default viewer?
a method to record metrics
batch import using standard identifiers would be helpful. ORCID has an interesting pattern for importing citations from CrossRef.
some sort of security policy that can be attached to the object or datastreams

bradspry commented 8 years ago

It is a repository, so I recommend focusing on ensuring repository objects are safe and sound.

premis
checksum
checksum checker
bag-it
fits

DonRichards commented 7 years ago

Accessibility Compliance Google Scholar Integration
Tracking (kinda like Google Analytics but on object level)

BuddyinKC commented 7 years ago

A default LICENSE datastream to store hidden pdf or xml license agreements with the work object with an integrated ingest mechanism to enable users to ingest a work object and a license object seamlessly would be very useful.

bradspry commented 7 years ago

@BuddyinKC, do you know the names of any machine-readable licenses or rights statements? I know of ONIX-PL, but ONIX-PL doesn't address rights. For example, are there any rights.xml schema which addresses IP Filtering rights, a.k.a. locking an object to a campus' IP range? Or a date-based license.xml or rights.xml like "we grant you streaming rights until MM-DD-YY"?

BuddyinKC commented 7 years ago

@bradspry, good question. A couple of things come to mind. There are ones for creative commons licenses. Also, A NISO working group has proposed best practices for access and license indicators (http://www.niso.org/apps/group_public/download.php/14223/rp-22-2015_ALI.pdf) that includes some XML implementation guidelines. There is some push to make this consistent across scholarly publishers and so there may be an opportunity to provide this consistency in IRs as well.

alexkent0 commented 7 years ago

Someway to include default copyright statements and the ability to select more? This could at least give people a place to start. http://rightsstatements.org/en/ could be a place to look for statements that could be applicable for a lot of sites.