I am building a search application for a website with solr + lucene.
There are some special keywords in the content,such as 'TF-708',which is processed by lucene as 'tf' and '708',
so if type keyword 'TF-708',the query is split into 'tf' or '708',which boost document with more 'TF' higher than less 'TF-708',
but we actually need 'TF-708' higher than others.
which one has good ideas to resolve this issue?