Sitecore Azure Search item level boosting

Abstract

You use search result boosting to score some topics higher than others in an index.

You can use the Sitecore Azure Search provider to tweak your results document score by boosting specific items so that they rank higher than others in search results. The results document score is dependent on the initial score, (which Azure calculates). In the item configuration area, you can boost items by defining their values so they rank higher than others. You can also define the scoring profile configuration and interpolation type in Sitecore configuration files.

You can configure boosting for Sitecore in the Content Editor, under the View tab, by selecting the Standard fields check box. Under the Indexing tab, in the Boost Value field, you can specify the value of an item, (from 1 to 10). To display certain items higher than others, assign them a higher boost value.

You can define the Scoring profile configuration, by navigating to App_Config\Sitecore\ContentSearch.Azure\ and opening the Sitecore.ContentSearch.Azure.DefaultIndexConfiguration.config file.

Sitecore supports several scoring profiles. However, only the profile that is currently active is defined in the <activeScoringProfileName> node. To apply a scoring profile to a specific index, you must add a <scoringProfilesRepository> node to that index. For example:

<configuration ref="contentSearch/indexConfigurations/defaultCloudIndexConfiguration">              <scoringProfilesRepository ref="contentSearch/scoringProfilesRepository"/>
</configuration>

When you add the node to any index, Sitecore automatically applies the scoring profile to that index. The scoring profile is included in the core, master, and web indexes by default. There is more information about scoring options in the file: Sitecore.ContentSearch.Azure.DefaultIndexConfiguration.config, under the <scoringProfilesRepository> node.

If you have two documents that have different Azure Scores for the same search query, then Sitecore displays the document with the highest score higher in the search results. In the following example Document 1 returns higher in the search results because it has the higher score, but when you apply boosting to the boost value, Document 2 ranks higher in the search results.

Document

1

2

Search score (initial)

10

5

Boost value

1

10

Search score (after boosting)

10

50

Note

To the increase the chance that your score will be higher and the document will display higher in search results, set the value in the Boost value field as close to the highest boost range point as possible. To affect the order of the documents, set a greater difference between the boost values of each item.