I am trying to use phonetic algorithm to perform (approx) search but I need some help on finalizing the algorithm since each algorithm has its pros and cons.
For Ex: Most of the phonetic algorithms matches 'tattoo' for the keyword 'Toyota'. Some fail to match hedison when searched for Hudson.. I am not expecting any particular result set when doing a search with phonetics but I don't want to return a completely different record..I am trying to make the results more relevant to what user is trying to search for since we don't have spell check implemented now.. It's ok if it doesnt return hedison for a search on Hudson but I don't want SOLR to return Tattoo when doing search for Toyota.. I am thinking of using 2 different phonetics algorithm (in 2 different fields) and filter the records based on both the fields... Is there a best practice for configuring phonetic filters? I don't want to re-invent the wheel if theres already a way to improve the accuracy.. -- View this message in context: http://lucene.472066.n3.nabble.com/Best-phonetic-configuration-tp4082671.html Sent from the Solr - User mailing list archive at Nabble.com.
