Ranking Tab

Site Search > Relevance > Relevance Modeling > Ranking Tab

The Ranking Tab of Site Search’s Model Details screen lets us apply “boost” multipliers to relevance scores of matches that fall within specific fields. You can also calculate a partial relevance score using a function, and then boost that value as desired. 

Note that you can turn the feature on and off using a checkbox. 

To apply a boost to a field, select the Search Field option and select the field from the drop-down list. Click the (+) button to create a slider control for that field. 

To calculate a partial relevance score based on the value of a field, select the Function option and type in a Solr Function Query expression. Click the (+) button to create a slider control.

Slide the controls left and right to establish the relative importance of a match in each field. A default (unboosted) field has a boost multiplier of 1. 

Some example functions you can use are:

  1. To boost if a field say ‘sitename’ contains a particular value (say Public) – termfreq(sitename,'Public')
  2. To boost by a date field – say ‘publish_date’ – recip(ms(NOW, publish_date),3.16e-11,10,1)

Use the Save Draft button to save your changes. The Publish button makes this Relevance Model available to the Relevance Preview screen.