Relevance ranking modules are the main components of the relevance ranking strategies for both record searches and dimension searches.
For information about how to use relevance ranking modules in relevance ranking strategies, see Ranking keyword search results.
The following table lists the relevance ranking modules and describes the sorting criteria that each specifies. For detailed information about relevance ranking modules, refer to Oracle Commerce Guided Search MDEX Engine Developer's Guide.
|
Relevance Ranking Module |
Description |
|
field |
Rank documents based on the search interface field with the highest priority in which it matched. By default, matches caused by cross-field matching are assigned a score of zero. The score for cross-field matches can be set explicitly by moving the <<CROSS_FIELD>> indicator up or down in the Selected Members list of the Search Interface editor. |
|
maxfield |
Behaves identically to the Field module, except in how it scores cross-field matches. Unlike Field, which assigns a static score to cross-field matches, Maximum Field selects the score of the highest-ranked field that contributed to the match. For example, if Record A and Record B each match several different members of a search interface, and Record B matches a member with a higher priority than any of the members matched by Record A, then Record B appears before Record A in results lists. |
|
glom |
Rank single-field matches above cross-field matches. |
|
nterms |
Rank based on the number of terms found. |
|
numfields |
Rank on the number of fields that completely match the query. |
|
phrase |
Rank on phrase, substring, or other; configurable. |
|
exact |
Rank on exact phrase match. Exact matches rank the highest. |
|
static |
Rank on specified sort key; configurable for sort key. |
|
first |
Rank based on how close the match is to the start of the field. |
|
proximity |
Rank based on how close the multi-terms are to each other in the matches. |
|
freq |
Rank based on how frequently the terms appear in the matched data. |
|
wfreq |
Same as freq but more weight given to information content. |
|
interp |
Weight by whether interpretation (for example, spelling) is used to get match. |
|
spell |
Ranks matches that have not been spell corrected above matches that have been spell corrected. |
|
stratify |
Rank records by stratifying them into groups defined in EQL expressions; used to boost or bury records in the result set. |
|
stem |
The stem module ranks matches produced by stemming below matches that are not produced by stemming. Stem assigns a rank of zero to matches from stemming, and a rank of one from all other sources. It ignores all other sorts of query expansion. |
|
thesaurus |
Rank matches due to thesaurus entries below other sorts of matches, |
The modules are applied in the order in which they are listed in the
relRankStrategy property in
assembler-context.xml.
For example, suppose that you configure a relevance ranking strategy that includes the modules Exact, Glom, and Static, as follows:
<property name="relRankStrategy" value="exact, glom, static(quantity_sold, descending)">
The records in the result list are sorted as follows:
