Tag: discovery layer

BagIt Profiles – @ruebot

• directory of data
• bag has what you’re bagging, data, contact email/name, organization information, profile identifier (JSON via a URI)
• pull in all the field values
• validate
• wrote a spec and send it to digital curation community
• can look up profiles in the registry

Okay, I got a little lost, but you can see more on github.

Internet Archive Torrent Collections (iaTorrent) – @ruebot

• see demo

Bookfinder – @TheRealArty & Steven

• I will write this up later probably as a separate blog post, or maybe journal article

TPL’s Web Services Architecture: Understanding the Big Picture – @waharnum

• many different systems that don’t easily communicate, which needs specialized knowledge even to do basic tasks
• address the challenges by translation, simplication, standardization
• Three tiers: Front End Systems (requests to back end) / TPL Web Services (REST) / Back End Systems (responds to front end)
• Example: TPL Website -> Account Web Services -> Symphony Web Services (Symphony) – and back
• can add new features and functions
• helps to solve the challenges mentioned
• also helps with reusability e.g. in addition to website, build mobile-friendly website, iPhone App
• Might end up with:
  • Front End (Website, mobile, App)
  • Middle Tier (Account Web Services, ebook Web Services, online payment web services)
  • Back End (symphony, overdrive, payment gateway, accounting systems)
• other benefits:
  • increase ease of knowledge transfer about how our systems work
  • follow modern best practice approach to building interoperating systems
  • reduce cost and integration time
• reduce learning time for new staff or consultants
• metrics: wish had resources
• bolting together a lot of things, not using a lot of custom code

Ladder (aka MyTPL 2) – @mjsuhonos

• wanted to solve problem: discovery layers suck
• problems:
  • not scalable
  • inflexible
  • read-only
  • expensive
• goals:
  • better than open source options (VuFind, Blacklight)
  • cheaper (than proprietary)
• scalable as WorldCat
• design:
  • schema-free/multi-schema (e.g. Dublin Core)
  • horizontally scalable (multi-node)
  • modern OSS components
• simple data model (RDF)
• Features:
  • hierarchical relations
  • clustering/de-duplication
  • versioning
  • real-time import & indexing
  • multi-thread/process
  • responsive UI
  • fully multilingual (18/10)
  • dynamic faceting
  • dynamic mapping modification
  • digital content storage (coming soon)
• built on a linked data
• not a discovery layer; it’s an integration platform

Heritage U of T – @ajmcalorum

• News Announcement and Promotional Video
• previously not centralized: hard drives, flickr, etc.
• need central repository for tri-campus initiative with search & discovery, preservation, long-term access to content and metadata, support for multiple formats (e.g. images, books, documents, video, exhibits)
• Drupal + Solr (search) + Fedora Commons (collection management, batch ingesting, metadata crosswalk, digital preservation) == islandora (digital asset management system)
• pilot: 8 parent collections (by format, by campus)
• exhibits in Drupal, not through islandora/fedora commons
• modules: internet archive book reader (OCR on the fly), galleria, colorbox
• official launch: 2 weeks ago

That’s it! Food and drinks time!