The Items for Query recommender is the primary algorithm that supports recommendations on the search results page. The items-for-query recommendations are based on how other users interacted with search results from the same query. This is one type of collaborative recommendation. For example, if users have searched for “titanic” in the past and many of them clicked on the search result for the DVD of the movie Titanic (as opposed to books or memorabilia), then this method boosts the DVD item on subsequent searches for “titanic”. Below is an example of items-for-query recommendations based on a search for “madonna”: Items-for-query examples To use this type of boosting, signals must be enabled but recommendations need not be.

​
Job configuration

The items-for-query recommendations:
  • Are implemented as a Query-to-Query Similarity job.
  • Use the output obtained from regular click signal aggregations from the job that provide query time signal boosts.
  • Job is an Alternating Least Squares (ALS) job like items-for-item recommendations and items-for-user recommendations, and requires system resources and time to execute.
    You can set more a more aggressive filtering strategy to decrease the run time, which typically translates into better recommendations. However, this strategy does not include coverage of as many queries as you may need. In that case, you might have to revert to more inclusive filters, which increases the run time.

​
Query pipeline configuration

The Boost with Signals query stage is part of the default query pipeline, so usually you do not need to add it. It takes its input from the aggregated signals collection (COLLECTION_NAME_signals_aggr) at query time.