#include <SimpleKLRetMethod.hpp>
Inheritance diagram for lemur::retrieval::SimpleKLScoreFunc:
Public Member Functions | |
void | setScoreMethod (enum SimpleKLParameter::adjustedScoreMethods adj) |
virtual double | matchedTermWeight (const lemur::api::QueryTerm *qTerm, const lemur::api::TextQueryRep *qRep, const lemur::api::DocInfo *info, const lemur::api::DocumentRep *dRep) const |
virtual double | adjustedScore (double origScore, const lemur::api::TextQueryRep *qRep, const lemur::api::DocumentRep *dRep) const |
score adjustment (e.g., appropriate length normalization) | |
Public Attributes | |
enum SimpleKLParameter::adjustedScoreMethods | adjScoreMethod |
The KL-divergence formula D(model_q || model_d), when used for ranking documents, can be computed efficiently by re-writing the formula as a sum over all matched terms in a query and a document. The details of such rewriting are described in the following two papers:
|
score adjustment (e.g., appropriate length normalization) The following are three different options for scoring ==== Option 1: query likelihood ============== ==== Option 2: cross-entropy (normalized query likelihood) ==== ==== Option 3: negative KL-divergence ==== |
|
|
|
|
|
|