We are attempting to find terms (actually words) that most frequently appear in some proximity to a given term.
When using facetting like this:
?q=myfield:tree&facet=on&facet.query=tree&facet.field=myfield it appears we are getting the counts for the most-occuring words in
myfield, next to
Question: However, we would like to kind of sub-facette the query, including the distances for each term (or an average distance).
myfield-field is stored, tokenized, the termvectors are stored and offset and position are stored with the termvector.