The shingle filter will likely be part of the strategy: https://cwiki.apache.org/confluence/display/solr/Filter+Descriptions#FilterDescriptions-ShingleFilter
and the keep word filter will help if you have an authority list: https://cwiki.apache.org/confluence/display/solr/Filter+Descriptions#FilterDescriptions-KeepWordFilter Joel Bernstein http://joelsolr.blogspot.com/ On Sun, Apr 9, 2017 at 1:54 AM, marotosg <marot...@gmail.com> wrote: > Hi, > > I am trying to extract locations from a location field which contains > location information in different formats. My initial idea is to extract > only UK and USA location and get them Standard. > > For instance if my field contains "Wakefield" then I will convert it to > "Wakefield West Yorkshire". > After that my new field will be a facet field with standard locations. > > I am trying to achieve this with the new SynonymGraphFilterFactory but waht > I get after applying it is a list of tokens like that. > Wakefield (1) > West (1) > Yorkshire (1). > > How is it possible to get all the tokens as one Wakefield West > Yorkshire(1). > > Here is the analyzer field definition > <analyzer type="index"> > <tokenizer class="solr.StandardTokenizerFactory"/> > <filter class="solr.SynonymGraphFilterFactory" > synonyms="locations-classifier.txt" ignoreCase="true" expand="false" > tokenizerFactory="solr.StandardTokenizerFactory"/> > <filter class="solr.FlattenGraphFilterFactory"/> > </analyzer> > > > thanks > > > > -- > View this message in context: http://lucene.472066.n3. > nabble.com/Tagging-Locations-using-SynonymGraphFilterFactory- > tp4329089.html > Sent from the Solr - User mailing list archive at Nabble.com. >
