search

ucsdlib / damspas-rd

A Digital Collections application based on Hyrax
MIT License
3 stars 2 forks source link

Question/multi-lingual search and display in DAMS5 #131

Open ghost opened 7 years ago

ghost commented 7 years ago

Descriptive summary

In Hyrax, is it possible to use a language attribute with prefLabel and altLabel in Concept, Agent, Place so they can be returned in search results and displayed in the UI?

Do we need to alter the D5 data model to make this possible?

Rationale

Users might only know a Concept, Agent, or Place in a particular language and will search for that. Even if there is an English label, the label in another language might be more common in a particular researcher discipline. Users who are not proficient in English would benefit if terms were displayed in languages other than English. This might also help identification and understanding of descriptive metadata (there can be subtle differences in meaning)

Expected behavior

Not sure, but... if a match is found on an a prefLabel or altLabel, in any language besides English, display the label that was searched for so the user understands why they got a match (maybe display in English also?)

mcritchlow commented 7 years ago

This "should" be possible. We'll need to look and see how to do this in the hyrax stack, but i'm sure the underlying RDF functionality for Ruby we're using supports it. The question I'd have is how easily accessible/renderable the data would be for views (such as forms).

So I think we could do a couple things if we want to pursue this more (which I agree is a great idea):

  1. R&D by dev team. See what's possible with adding multi-lingual instances of things like labels and prefLables. That could be in this ticket, potentially.
  2. More specifically document where we'll want to use this. Just authority records, or more globally?
  3. More specifically document the expected associated behaviors in the application. indexing (as Greg noted) as well as UI (views, forms, etc.)
  4. Related to 2+3, does the multi-lingual data need to be persisted in Fedora, or is this something that purely a search/discovery concern and related to external authority records where the data already exists? If so, this could potentially be more directly targeted to a custom indexing strategy for external authority records, where we de-reference and index this metadata into Solr only.

Not sure if this would need to go to Products or not. It's an enhancement of sorts, for sure, but seems like a very clear use of linked data we'd want to leverage. I defer to @gamontoya

gamontoya commented 6 years ago

Revisit in future.