Hi Solr Comunity,
My use case is the following:
Users make free text searches and hardly use the filters for an e-commerce
site. The idea is that we build a content based profile, simply aggregating
all purchases before. Cretae map where the keys are terms from the
aggregated product descriptions and the values are frequencies.
When the customer searches we apply reranking where the reranking query
would be a "weighted top 20 terms". I found in the documentation the
following syntax:
q=greetings&rq={!rerank reRankQuery=$rqq reRankDocs=1000
reRankWeight=3}&rqq=(hi+hello+hey+hiya)
1. How can I boost terms in the reRankQuery as I have different
weights and importance based on the customer profile?
2. Can I use edismax query parser as a rerank parser?
3. If reranking does not support boosting by terms, what is the
proposed alternative?
Thanks in advance,
Roland
