Learn-to-rank settings

The SmartHub learn-to-rank component enables SmartHub to use clustering algorithms to build clusters of similar queries and the documents that are most associated or interacted with after running a query. A cluster contains the following:

• Queries that contain similar keywords.

• Documents that were interacted with after running the query.

• Other queries that lead to the same documents, without necessarily containing similar keywords.

The Learn-To-Rank component is implemented on top of SmartAnalytics data to support four main features:

• Results boosting: The component will boost documents from a cluster if the search query is similar to other queries in that cluster.

• User segmentation boosting: The component will boost and suggest documents based on interaction between users with similar classifications, such as role, location, job type, etc.

• Query suggestions: The component will provide suggestions as query text is entered into the search box.

• Content suggestions: The component will suggest similar results to the cluster, excluding those currently on the page.

The learn-to-rank settings page allows you to configure specific controls and values for the learn-to-rank component. For more information, see how to use learn-to-rank features.