I am not sure if a Query Filter would reduce the retrieval time in the following scenario. The idea is to retrieve top N documents where the query terms appear near one another within a document.
Let A = Total number of query terms B = Subset of A where Number of Query terms(B) << Number of Query terms(A) C = Pairs of terms in B But I make sure that the pair selection results in, the Number of term pairs(C) = Number of query terms(B) I need to retrieve the documents by formulating the query as BooleanQuery(TermQuery(A) OR SpanNearQuery(C)) Option 1: BooleanQuery(TermQuery(A) OR SpanNearQuery(C)) Option 2: Step1: BooleanQuery(TermQuery(A)) Step2: Use a query filter top N documents. The number of documents N is much less compared to the entire document collection. Step3: SpanNearQuery(C) on those top N documents Since Span queries are slower. Considering the fact that the number of documents N would be much less than the total number of documents in the index, is it better to query them on reduced number of N documents as in Option 2? Thanks, Rajesh Munavalli
