It's solved by "publishing" it in an open access repository - so the data is available to public, but all the ratings and citations and relevant rankings go to the closed access journal where you initially published it.