Additional or Quality Metadata for Google Scholar (aka, The Scholar)

[mtribone]

We need to rework the metadata per the recommendations from Google in regards to scholarly literature.

• https://scholar.google.com/intl/en/scholar/inclusion.html#indexing
• Place each article and each abstract in a separate HTML or PDF file
• Use Dublin Core tags (e.g., DC.title) as a last resort
• Only scholarly papers are appropriate for inclusion in Google Scholar; each paper needs to be listed on a separate URL; and at least the full author-written abstract must be clearly visible on the URL that you wish to be included in Google Scholar search results.

Example Work that we would want to be indexed
https://scholarsphere.psu.edu/concern/generic_works/xwd375x69k

Datasets should be found in Google Dataset Search.
https://developers.google.com/search/docs/data-types/dataset

[DanCoughlin]

There are a couple of things that would be very helpful for indexing if possible:

• Adding the publication date to the citation_publication_date tags, so they aren't blank, e.g.
  A common repository indexing error is using the upload/online date; only the formal publication date should be entered in this field.
• Adding only the title of the item, and not the file name, to the citation_title, e.g.
  view-source:https://scholarsphere.psu.edu/files/rf55z7756

looks great.

A file with .csv at the end, which would be a red flag for the indexing system. It would not be considered a publication

It's better not to add empty metatags. If you don't have the publication date of an item, for example, it's best not to include the citation_publication_date tag.

Additionally, the journal specific information like vol/issue and page numbers can be left off for repository items. The really crucial items for ScholarSphere would be:

• citation_title
• citation_author
• citation_publication_date
• citation_pdf_url

[mtribone]

Placeholder for questions to help solve the indexing issues

1. Does each article require the abstract to be in a separate HTML or PDF file?
2. What file formats are excepted? Seems like preservation formats are not the required file formats?

[mtribone]

Google Scholar view Partial
https://github.com/psu-stewardship/scholarsphere/blob/2735de0c3994d2f0e32ac2ef5f5219af528f83d8/app/views/shared/_gscholar.html.erb

[mtribone]

Perhaps we should require the publication date if a work has the resource type of article, book, journal, part of book, research paper? This would require a change to the new/edit work form to pull the publication date out of the additional metadata. Or we could remove the citation_publication_date from the meta for Google Scholar if it is blank. Might be better to get a date.

[mtribone]

We will also need to redo Batch Create because the process uses the filenames as the title of the work. It does not remove the file extension from the title.