com.bericotech.clavin.resolver.multipart

Class DefaultScorer

java.lang.Object

com.bericotech.clavin.resolver.multipart.DefaultScorer

All Implemented Interfaces:

Scorer

public class DefaultScorer
extends Object
implements Scorer

The default scorer combines several weighted metrics to return a score in the range [0.0, 1.0]. The score components are: - Match Ratio (M) the number of matched terms / the number of input terms - Component Score (C) A metric boosting the scores of matches containing particular location types. From highest to lowest: CITY, COUNTRY, ADMIN1, ADMIN2, ADMIN3, ADMIN4 - Average Damerau-Levenshtein (DL) Distance (D) The average DL distance (number of edits) between the original search term and the matched name. This metric is inverted (1/DL) because lower DL numbers are better. Identical matches are given a score of 1. - Average Search Depth (S) The average rank within the search results for the parents of the most specific match. Matches including only a single location will receive a score of 1. The components are weighted according to the following formula: (0.40 * M) + (0.25 * C) + (0.20 * D) + (0.15 * S)

Constructor Summary

Constructors

Constructor and Description
DefaultScorer()

Method Summary

Methods

Modifier and Type	Method and Description
double	getMaximumScore() Get the maximum score returned by this scorer.
double	getMinimumScore() Get the minimum score returned by this scorer.
double	score(List<String> terms, MatchedLocation candidate) Generate a score representing how well the candidate result matches the search terms.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

DefaultScorer

public DefaultScorer()

Method Detail

score

public double score(List<String> terms,
           MatchedLocation candidate)

Description copied from interface: Scorer

Generate a score representing how well the candidate result matches the search terms. Higher scores indicate better matches.

Specified by:

score in interface Scorer

Parameters:

terms - the original search terms

candidate - the candidate match

Returns:

a numeric score indicating the quality of this candidate, higher scores are better

getMinimumScore

public double getMinimumScore()

Description copied from interface: Scorer

Get the minimum score returned by this scorer.

Specified by:

getMinimumScore in interface Scorer

Returns:

the minimum score for a candidate

getMaximumScore

public double getMaximumScore()

Description copied from interface: Scorer

Get the maximum score returned by this scorer.

Specified by:

getMaximumScore in interface Scorer

Returns:

the maximum score for a candidate