Virtual Agent Ai search Results

Rivka Br
Tera Contributor

Hi,

I want add to the search results in ai search in virtual agent the score
like in the search preview
in the topic of ai search (try to add it to evam with score but not return)
like it:

RivkaBr_0-1708605620579.png

 

2 REPLIES 2

Community Alums
Not applicable

Hi @Rivka Br ,

Configure a result improvement rule to increase the relevancy scores of search results that satisfy a custom condition.

Before you begin

Create your result improvement rule first to define your trigger conditions, then follow these steps to define the action taken when the rule activates. For steps to create a result improvement rule, see Create a result improvement rule.

Role required: ais_admin

About this task

When a result improvement rule with a boost action activates, AI Search modifies the relevancy scores for search results from an indexed source specified in the action. You can choose one of these boost types:
  • Boost all search results from the specified indexed source.
  • Boost only search results from the specified indexed source with a field value that matches a dynamically populated element of the searcher's user context.
  • Boost only search results from the specified indexed source with a field value that matches a static string value specified in the action.

You can specify a boost weight for each boost action. This is an integer value that AI Search uses to alter the relevancy scores for search results affected by your boost action. Boost weights can be positive or negative.

Positive boost weights increase the relevancy scores for boosted search results. Each boost weight of 1 increases a boosted search result's relevancy score by 0.1% of its original value. For example, if you set a boost weight of 100, you increase each boosted search result's relevancy by +10%, so a search result with an original relevancy score of 60 gets a final relevancy score of 66. Setting a boost weight of 1000 adds +100%, doubling the relevancy scores for boosted search results.

Negative boost weights reduce the relevancy scores for boosted search results. Each boost weight of –1 reduces a boosted result's relevancy score by 0.1% of its original value. As an example, if you set a boost weight of –150, you reduce a boosted search result's relevancy by 15%, so a search result with an original relevancy score of 100 gets a final relevancy score of 85. Any boost weight of –1000 or lower reduces a boosted search result's relevancy score to 0, causing it to appear at the end of the search result list.

When a single search result is affected by multiple boost actions, AI Search adds the individual boost weights together to determine the total relevancy score modifier. For example, a search result boosted by three rules with boost weights 100, 150, and –25 has a total boost weight of 225, increasing its final relevancy score by 22.5% of its original value.

Procedure

  1. In the Actions related list, select Create Boost Action.
  2. On the Rule - Action Mapping form, set the boost weight and fill in the other fields.
  3. Select Submit.
  4. (Optional) Edit the Order values for the result improvement rule's boost actions.

What to do next

To make new or updated result improvement rule settings take effect, publish the search profile you edited. For details on publishing a search profile, see Publish a search profile.

 

 

Hi,

I want to show the users when they search anything
from the virtual agent the results include the score