US Patent 11580153 Lookalike expansion of source-based custom audience by an online system

An online system generates a cluster group and uses membership in the cluster group as an eligibility criteria for presenting a content item. The online system receives a request from a third party system to present the content item. The online system also receives identification information about users who have visited webpages associated with the third party system and descriptive information associated with the webpages. Based on the descriptive information, the online system extracts tags for the webpages and classifies the webpages into one or more categories that include a category associated with the content item. The online system generates a seed group that includes users who visited webpages in the category associated with the content item. The online system further expands the seed group to a cluster group by applying a cluster model to one or more characteristics of each candidate user not included in the seed group.