Is it possible to write an MLT query that assigns higher "priority" to terms that are rare across all documents than to terms that are frequent on the specific document? From what I could gather the boost_terms parameter only allows me to boost specific terms, as opposed to boosting rare terms.
Thanks!