|
Qizx/open 4.1 API | |||||||||
| PREV CLASS NEXT CLASS | FRAMES NO FRAMES | |||||||||
| SUMMARY: NESTED | FIELD | CONSTR | METHOD | DETAIL: FIELD | CONSTR | METHOD | |||||||||
public interface Scorer
Customizable part of the full-text scoring algorithm.
This interface defines:
| Field Summary | |
|---|---|
static float |
CORE_SCORE
Score returned when the corresponding document does not belong to an XML Library. |
| Method Summary | |
|---|---|
float |
normalizeScore(float rawScore)
Normalizes a score so that its value is between 0 and 1 |
float |
normAll(float[] subWeights)
Computes the weight norm associated with a conjunction ('ftand' or 'all') of full-text selections. |
float |
normOr(float[] subWeights)
Computes the weight norm associated with a disjunction ('ftor' or 'any') of full-text selections. |
float |
normWord(float inverseDocFrequency)
Computes the weight associated with a simple word. |
float |
scoreAll(float[] subScores)
Computes the score of a conjunction ('ftand' or 'all') of full-text selections. |
float |
scoreOr(float[] subScores,
int scoreCount)
Computes the score of a disjunction ('ftor' or 'any') of full-text selections. |
float |
scoreWord(float norm,
float termFrequency)
Computes the score of a single word. |
| Field Detail |
|---|
static final float CORE_SCORE
| Method Detail |
|---|
float normWord(float inverseDocFrequency)
inverseDocFrequency - inverse of the fraction of documents that
contain this term
float normAll(float[] subWeights)
subWeights - weight values computed for sub-selections.
float normOr(float[] subWeights)
subWeights - weight values computed for sub-selections.
float scoreWord(float norm,
float termFrequency)
norm - normalized weight of word (as computed by normWord(float)).termFrequency - relative term frequency in the current document
(number of occurrences divided by average number of occurrences in all
documents).
float scoreAll(float[] subScores)
subScores - weight values computed for sub-selections.
float scoreOr(float[] subScores,
int scoreCount)
subScores - weight values computed for sub-selections.scoreCount - number of values in scores
float normalizeScore(float rawScore)
rawScore - a positive value possibly greater than 1
|
© 2010 Axyana Software | |||||||||
| PREV CLASS NEXT CLASS | FRAMES NO FRAMES | |||||||||
| SUMMARY: NESTED | FIELD | CONSTR | METHOD | DETAIL: FIELD | CONSTR | METHOD | |||||||||