Results > Manipulate Result Relevance > Use the BIAS Field Specifier to Boost Relevance

Use the BIAS Field Specifier to Boost Relevance

Topics in this Section

The BIAS field specifier allows you to bias the score of results at query time according to the numerical proximity of the specified field to a given value. It ignores initial $, £ or characters in the field.
Specify BIAS using the format:
BIAS{optimum,range,percentage}
where,
 
is a positive number that determines the range of the optimum. If the specified field contains a value that is in the range of (optimumrange) to (optimum + range), the result weight increases or decreases according to the specified percentage.
is a percentage in the range –100 to 100. If the value of the specified field is within the specified range, the score of the result increases or decreases according to how close the value is to the specified optimum.
For example:
n http://IDOLhost:port/action=Query&FieldText=BIAS{100,50,10}:*/PRICE
A document whose PRICE field value is within the range 50 either side of 100 has its weight increased on a linear scale from 10 percent if the price is 100, to zero percent if the price is 50 or 150:
n http://IDOLhost:port/action=Query&FieldText=BIAS{100,50,-10}:*/PRICE
A document whose PRICE field value is within the range 50 either side of 100 has its weight decreased on a linear scale from –10 percent if the price is 100, to –0 percent if the price is 50 or 150:
*
You can also use the BIAS field specifier to bias the score of results according to the numerical proximity in their autn_date metafield to a given value. For example:
FieldText=BIAS{1103918400,259200,25}:autn_date
A document whose autn_date field value is within the range 259200 either side of 1103918400 has its weight increased on a linear scale from 25 percent if the price is 1103918400, to zero percent, if the date is 1103659200 or 1104177600.
Related Topics 
*