Show Similar Works on works pages (from 368)

[gigamorph]

Problem: LUX should offer an option to show Works related based on elements shown in wireframe (add here)

Expected Behavior/Solution: LUX should offer 5-10 similar results on a Works, with the option of running a search to see all similar results. Expectation is that there may be millions of similar results.

Requirements: What critieria are used for similar people? How is "relevance" determined? Keywords from description statement, itemName, itemPrimaryName

• Keywords from description statement, itemName, itemPrimaryName
• Classification
• Material
• produced in a similar time period -- (+/- 10 years if from after 1900) -- (+/- 20 year if from after 1700) -- (+/- 35 years if from 1700 or earlier)

Needed for promotion:

• [ ] review similar searches results/requirements definition - Peter/Kelly
• [ ] placement of similar on Works page - Mike
• [ ] ~Review potential for Visual Image image similarity (AI matching)~
• [ ] Wireframe/mockup
• [ ] committee discussion
• [ ] feasibility/team discussion
• [ ] front end req
• [ ] back end req
• [ ] questions

UAT/LUX Examples:

• Example endpoints from LUX to use for development and testing, if applicable.

Dependencies/Blocks:

• ML Research Ticket: https://git.yale.edu/lux-its/marklogic/issues/519

Related Github Issues:

• https://github.com/project-lux/lux-frontend/issues/5 (objects)
• https://github.com/project-lux/lux-frontend/issues/202 (people & groups)
• https://github.com/project-lux/lux-marklogic/issues/17
• https://github.com/project-lux/lux-marklogic/issues/18

Related links:

Wireframe: Old wireframe that may need work: https://nj46vq.axshare.com/?id=kc0vxl&p=similar&g=1 (from original ticket convo)

[roamye]

Is there something we can do for this to be moved along? or is this more for the future? One of the first req was for Kelly and Heather to build req for what constitutes as "similar" for each attribute. Would this still be our next step?

Same questions for the related issue: #5

@prowns @miketullo95 @kkdavis14

[roamye]

Bring to Core Team meeting. Is there a user need for this? Should we do some spot testing to see what kind of results we get from the current ML similar parameters?

@clarkepeterf and @kkdavis14 - would be great to have some examples shared here.

[prowns]

Core team: this is valuable; continue to prep