search

project-lux / lux-frontend

Web front end of LUX
Apache License 2.0
3 stars 0 forks source link

Expanding Search Query for Objects in Collection #221

Closed roamye closed 6 days ago

roamye commented 2 months ago

Problem Description: Under the group record Yale University Art Gallery, 1940 - the Objects in Collection Tab only shows three results. However, there should be more results listed as there are multiple objects under Yale University Art Gallery. (example)

Note: The results under YUAG is caused by how their data is vended. All other Units do not show as a collection because collections are handled by the individual departments.

Expected Behavior/Solution: Move forward with expanding the search query to include an additional block that can search for all objects in a specific collection.

~Depending on the IT meeting on 05/10, we will move forward with: (a) label change to accurately define the results in Objects in Collection or~ ~(b) have all objects under the Objects in Collection, such as Yale University Library 1701- does.~

Requirements:

Needed for promotion: If an item on the list is not needed, it should be crossed off but not removed.

UAT/LUX Examples: Set pages:

Group pages

Dependencies/Blocks:

Related Github Issues:

Related links:

Wireframe/Mockup:: image

kamerynB commented 1 month ago

@prowns @roamye We need a label to promote this to prioritization.

prowns commented 1 month ago

agenda for interactions

prowns commented 1 month ago

@kamerynB - what happened to the Objects in Collection Tab (see bugherd)? I was going to bring to interactions but it is gone.

kamerynB commented 1 month ago

@prowns The HAL link is not in the record. It is lux:departmentItems.

prowns commented 1 month ago

Not appearing for YUAG, YCBA or YPM. @kkdavis14 - is this a data issue?

"lux:departmentItems": { "href": "https://lux.collections.yale.edu/api/search/item?q=%7B%22memberOf%22%3A%7B%22curatedBy%22%3A%7B%22id%22%3A%22https%3A%2F%2Flux.collections.yale.edu%2Fdata%2Fgroup%2Fd07b9b86-0a1e-4026-aa4c-8ecba8bbd9c9%22%7D%7D%7D", "_estimate": 1

prowns commented 1 month ago

Additional context/analysis for #agenda.

The "tab" is not appearing on any of the UAT example pages (per Kam, a HAL link issue?), but I also don't understand what is missing from the YUAG data, per the note a the top.

We will see this tab a lot in the UI when users click on Collection in results snippets and get to the set. Group: https://lux.collections.yale.edu/view/group/b6cb6d82-3a56-443f-8826-139c201afd2a Set: https://lux.collections.yale.edu/view/set/812660f3-946a-4941-aa85-6ac4fffaa22a

Question: is there ever a need for a "Set" page for a Responsible Unit? If so, then how would a user get to that SET page? Would we add a Responsible unit link to the results snippet?

prowns commented 1 month ago

This will be solved by hierarchy searching/transitive query. Will need to change HAL link to use transitive query. Blocked by ML work in this area. Need to find that issue.

roamye commented 1 month ago

tix forming 6/17: Objects that are curated by this group or members of this group can be added to the query to make it bigger.

Do we want to expand the query? #agenda

Maybe the departments can be more obvious/noticeable in the top-level group. (potential solution)

@prowns cc: @clarkepeterf

brent-hartwig commented 1 month ago

@prowns, as I understand this ticket, we can either change a label (current scope) or have the query find all objects associated to a collection, regardless of the data going through a department. More generally, we're looking to start with a collection, hop a specific predicate until we reach the objects, regardless of how many times we have to hop. (And have that ability for other predicates / queries.)

Options that come to mind:

  1. CTS can partially meet this need. It would need to be finite number of hops, but that may suffice for the current dataset.
  2. Offer a way to execute select searches outside of CTS (e.g., SPARQL).
  3. Wait for the Optic/API exercise to complete. If Optic is selected, this will be a freebie.

We could select no. 1 or 2 in advance of no. 3, and switch when the time comes.

We should plan a performance test for any selected option.

clarkepeterf commented 3 weeks ago

I think the easiest option is to expand the search to look for objects either curated by this agent, or by a member of this agent:

https://lux-front-tst.collections.yale.edu/view/results/objects?q=%7B%22OR%22%3A%5B%7B%22memberOf%22%3A%7B%22curatedBy%22%3A%7B%22id%22%3A%22https%3A%2F%2Flux.collections.yale.edu%2Fdata%2Fgroup%2F6086b58d-941d-41e3-87c1-e00e96952ffb%22%7D%7D%7D%2C%7B%22memberOf%22%3A%7B%22curatedBy%22%3A%7B%22memberOf%22%3A%7B%22id%22%3A%22https%3A%2F%2Flux.collections.yale.edu%2Fdata%2Fgroup%2F6086b58d-941d-41e3-87c1-e00e96952ffb%22%7D%7D%7D%7D%5D%7D

roamye commented 3 weeks ago

From it meeting 6/21 this will be brought up to UAT to discuss which option to move forward with. Will update tix after meeting 6/27.

roamye commented 3 weeks ago

@brent-hartwig - could ML 11.3 help with option # 3 listed in your comment?

"Wait for the Optic/API exercise to complete. If Optic is selected, this will be a freebie."

roamye commented 3 weeks ago

From UAT meeting 6/26: It is OK to move forward with option 1 (expanding the search) as mentioned by @brent-hartwig @clarkepeterf while we await if option 3 is possible.

brent-hartwig commented 3 weeks ago

@brent-hartwig - could ML 11.3 help with option # 3 listed in your comment?

"Wait for the Optic/API exercise to complete. If Optic is selected, this will be a freebie."

@roamye, ML 11.3 includes an Optic optimization we would want should we switch to Optic. But Optic in previous ML versions offer the functionality required by this ticket.

roamye commented 6 days ago

Closing this ticket as it should be a MT tix.

That tix can be found here: https://github.com/project-lux/lux-middletier/issues/76 and has moved onto backlog.