code4lib – Page 4 – Learning (Lib)Tech

XML Databases and Document Stores, Michael Joyce, SFU

EXIST for storing XML
indexing and searhcing provided by Lucene
doesn’t include analysis, but can used Stanford named entity recognizer
MASHUP of over 200 books mapped using statistical tagger and dbpedia, great for getting people started
similarity matching to find similar content

built at SFU in migrate digital collections to Islandora
built in modular way, and flexible
doesn’t move to Islandora, just prepares content
can have contentdm or csv/excel, prepares batch packages that Islandroa expects
can run as many times as you want to get it right before loading to Islandora
despite the name, will let you use it in other contexts e.g. move to Hydra
parse into MODS or DC, which is what Islandora exists, but could be written to do others
post-write hook scripts, allows scripts to be run in the background
modules can be used e.g. metadata manipulator
need to write config file, but not more complex than most
metadata mapping file can also add metadata that wasn’t in incoming data
backend tool written in PHP
hope to get people get interested to contribute different toolchanges e.g. Dspace

BC ELN started Arca, IR for all post-secondary schools
migrated from ContentDM to Islandora
created custom displays to display custom metadata
had to manually regenerate derivatives, but would want to script it in the future
most issues left over has to do with display e.g. list vs. grid per collection

very little data is linked at this point
want to propose that authors have persistent digital identifier through ORCID to build profiles, use ORCID to markup and import citations
can add schema and do something similar to call number
show related articles currently based on search but if use ID, can build richer ways to show related articles e.g. same person, others that they’ve collaborated with
can build citation and reference lists

Provincial Digital Library, but didn’t really consult anybody.
put some money to BC ELN to build Arca
next step is to build discovery layer on top of Arca and other digital collections
had a meeting, will get report in ~2 weeks
larger symposium next to BCLA; and will likely be other smaller groups
the ministries have said there is no money to do this
come up with some solutions and they might be able to find money
a task group is forming to see what we might do
digital collections environmental scan done by Mike Conroy
the focus is to benefit the whole province

Time for a snack!