Prioritizing Full-text Hit Lists

Full-text hit lists are weighted by default criteria such as number and proximity of search terms and by date and number of file versions. Newer objects and objects with a greater number of versions have a higher weighting.

In addition, the weighting can be configured depending on the following criteria:

• Metadata fields: If the search term is found in a configured metadata field, then the object has a higher weighting.

• Objects with metadata: If the search term is found in an object that is indexed in an metadata field with the configured value, then the object has a higher weighting.

Prioritizations use significantly more system resources and therefore should only be used sparingly.

The configuration is carried out via the search-prod.yml file located in the \config directory of yuuvis® RAD service-manager.

The fields need to be specified with the technical names used by Elasticsearch for the configuration.

You determine the Elasticsearch name by doing the following:

• Call up the following address:
  http://<host-gateway>/rest-ws/service/system/type/list.json?elements=true

• In the schema shown, look for the field using the technical name (name), the display name (label), or the description (description) shown in yuuvis® RAD designer. For the configuration you need the name indexname specified in the same section.

Prioritization of Fields

The syntax for prioritizing hits in fields has the following structure:

score.fields:'field_indexname_1;field_indexname_2;…'

Fields for which the objects should be prioritized higher if the search term is found there are listed. The prioritization takes place from left to right, which means that the first field has the highest priority.

Example:

score.fields:'str_rechnung_kundennummer;str_rechnung_referenz_nummer'

A customer number, which can occur in a 'customer number' field as well as in a 'reference number' field, is searched for. Hits in the 'customer number' field are prioritized over hits in the 'reference number' field.

Prioritization of Objects

The syntax for prioritizing hits in fields with specified metadata has the following structure:

score.setting:'field_indexname:field_value_1-weight_1,field_value_2-weight_2,field_value_3-weight_3'

Fields with values and weightings are specified.

Example:

score.setting:'str_document_status:aktiv-50,offen-20,inaktiv-10'

Hits for which the field 'status' is indexed with 'active' are weighted higher than hits with the status 'open' and these higher than hits with the status 'inactive'. The 'weight' values can be any value and they are only proportionally related to each other.

Multiple fields with values of one object type can be specified for prioritization using metadata.

Example:

score setting:'key_document_status:aktiv-50;key_type:vertragsunterlagen-50,dokument-20'

Hits with the status 'active' are scored higher and hits with type 'contract documents' are scored higher than hits with type 'document'.