Clustering users of a social networking system based on user interactions with content items associated with a topic

Abstract

A social networking system presents users with a content items and ad requests, which may include targeting criteria specifying a topic. Interactions by users who were presented with an advertisement from an ad request including targeting criteria specifying the topic are stored by the social networking system and used to identify a cluster group of additional users having characteristics similar to characteristics of users who were presented with the advertisement from the ad request including targeting criteria specifying the topic and who interacted with the advertisement. The social networking system determines scores for additional users in the cluster group based on measures of similarity between the additional users and the users who were presented with the advertisement and who interacted with the advertisement. Based on the determined scores, the social networking system associates additional users in the cluster group with the topic.

1 - 20 . (canceled) 
     
     
         21 . A system, comprising:
 a processor;   a memory storing instructions, which when executed by the processor, cause the processor to:
 identify a cluster group of target users having a threshold measure of similarity to prior users who have previously interacted with prior digital content; 
 determine a likelihood that each of the target users in the cluster group will interact with new digital content by applying a machine learning model to the cluster group of target users based on the threshold measure of similarity; and 
 identify an opportunity to present new digital content to at least one of the target users based at least in part of the determined likelihood that the at least one of the target users will interact with the new digital content, the new digital content having similar targeting criteria as the prior digital content. 
   
     
     
         22 . The system of  claim 21 , wherein the cluster group is identified based on:
 identifying at least one characteristic in the prior digital content with which prior users have previously interacted.   identifying a pool of users that share the at least one characteristic with the prior users;   selecting, from the pool of users, additional target users to add to the cluster group, based on a threshold value, wherein the threshold value comprises at least one of a number characteristics or a percentage of characteristics shared between the pool of users and the prior users; and   forming the cluster group of target users based on at least the selected additional target users.   
     
     
         23 . The system of  claim 22 , wherein selecting the additional target users from the pool of users further comprises:
 ranking the additional target users in the cluster group based at least in part the threshold number or the threshold percentage of the at least one characteristic that matches the pool of users with the prior users.   
     
     
         24 . The system of  claim 22 , wherein applying the machine learning model further comprises:
 training the machine learning model using information associated with the at least one identified characteristic, the threshold measure of similarity, or selected potential target users, wherein the machine learning model comprises at least one set of model parameters determined based at least in part on tendencies of the cluster group of target users.   
     
     
         25 . The system of  claim 21 , wherein determining a likelihood that each of the target users in the cluster group will interact with new digital content comprises:
 calculating a value for each of the target users in the cluster group using the machine learning model, wherein the value represents the likelihood that each of the target users in the cluster group will interact with new digital content, and the value is calculated based at least in part on the threshold measure of a similarity between the target users in the cluster group and the prior users or information associated with prior users who have previously interacted with prior digital content.   
     
     
         26 . The system of  claim 25 , wherein the value is calculated based at least in part on the threshold measure of a similarity between the target users in the cluster group and the prior users or information associated with prior users who have previously interacted with prior digital content. 
     
     
         27 . The system of  claim 25 , wherein the value is calculated based at least in part on information associated with prior users who have previously interacted with prior digital content. 
     
     
         28 . The system of  claim 21 , wherein the opportunity to present new digital content is further based on one or more digital content criteria, wherein the one or more digital content criteria is determined by at least one of a digital content provider, a digital content host, or a user. 
     
     
         29 . The system of  claim 21 , wherein identifying the opportunity to present new digital content further comprises:
 ranking the target users in the cluster group based at least in part on the determined likelihood that the at least one of the target users will interact with the new digital content.   
     
     
         30 . The system of  claim 21 , further comprising:
 presenting the new digital content to a mobile device associated with at least one of the target users in the cluster group based on the identified opportunity.   
     
     
         31 . A method, comprising:
 identifying, by a processor of a computing system, a cluster group of target users having a threshold measure of similarity to prior users who have previously interacted with prior digital content;   determining a likelihood that each of the target users in the cluster group will interact with new digital content by applying a machine learning model to the cluster group of target users based on the threshold measure of similarity; and   identifying an opportunity to present new digital content to at least one of the target users based at least in part of the determined likelihood that the at least one of the target users will interact with the new digital content, the new digital content having similar targeting criteria as the prior digital content.   
     
     
         32 . The method of  claim 31 , wherein the cluster group is identified based on:
 identifying at least one characteristic in the prior digital content with which prior users have previously interacted.   identifying a pool of users that share the at least one characteristic with the prior users;   selecting, from the pool of users, additional target users to add to the cluster group, based on a threshold value, wherein the threshold value comprises at least one of a number characteristics or a percentage of characteristics shared between the pool of users and the prior users; and   forming the cluster group of target users based on at least the selected additional target users.   
     
     
         33 . The method of  claim 31 , wherein determining a likelihood that each of the target users in the cluster group will interact with new digital content comprises:
 calculating a value for each of the target users in the cluster group using the machine learning model,
 wherein the value represents the likelihood that each of the target users in the cluster group will interact with new digital content, and the value is calculated based at least in part on the threshold measure of a similarity between the target users in the cluster group and the prior users or information associated with prior users who have previously interacted with prior digital content, and 
 wherein the value is calculated based at least in part on the threshold measure of a similarity between the target users in the cluster group and the prior users or information associated with prior users who have previously interacted with prior digital content. 
   
     
     
         34 . The method of  claim 31 , wherein the opportunity to present new digital content is further based on one or more digital content criteria, wherein the one or more digital content criteria is determined by at least one of a digital content provider, a digital content host, or a user. 
     
     
         35 . The method of  claim 31 , wherein identifying the opportunity to present new digital content further comprises:
 ranking the target users in the cluster group based at least in part on the determined likelihood that the at least one of the target users will interact with the new digital content.   
     
     
         36 . A non-transitory computer-readable storage medium having an executable stored thereon, which when executed instructs a processor to, comprising:
 identifying, by a processor of a computing system, a cluster group of target users having a threshold measure of similarity to prior users who have previously interacted with prior digital content;   determining a likelihood that each of the target users in the cluster group will interact with new digital content by applying a machine learning model to the cluster group of target users based on the threshold measure of similarity; and   identifying an opportunity to present new digital content to at least one of the target users based at least in part of the determined likelihood that the at least one of the target users will interact with the new digital content, the new digital content having similar targeting criteria as the prior digital content.   
     
     
         37 . The non-transitory computer-readable storage medium of  claim 36 , wherein the cluster group is identified based on:
 identifying at least one characteristic in the prior digital content with which prior users have previously interacted.   identifying a pool of users that share the at least one characteristic with the prior users;   selecting, from the pool of users, additional target users to add to the cluster group, based on a threshold value, wherein the threshold value comprises at least one of a number characteristics or a percentage of characteristics shared between the pool of users and the prior users; and   forming the cluster group of target users based on at least the selected additional target users.   
     
     
         38 . The non-transitory computer-readable storage medium of  claim 36 , wherein determining a likelihood that each of the target users in the cluster group will interact with new digital content comprises:
 calculating a value for each of the target users in the cluster group using the machine learning model,
 wherein the value represents the likelihood that each of the target users in the cluster group will interact with new digital content, and the value is calculated based at least in part on the threshold measure of a similarity between the target users in the cluster group and the prior users or information associated with prior users who have previously interacted with prior digital content, and 
 wherein the value is calculated based at least in part on the threshold measure of a similarity between the target users in the cluster group and the prior users or information associated with prior users who have previously interacted with prior digital content. 
   
     
     
         39 . The non-transitory computer-readable storage medium of  claim 36 , wherein the opportunity to present new digital content is further based on one or more digital content criteria, wherein the one or more digital content criteria is determined by at least one of a digital content provider, a digital content host, or a user. 
     
     
         40 . The non-transitory computer-readable storage medium of  claim 36 , wherein identifying the opportunity to present new digital content further comprises:
 ranking the target users in the cluster group based at least in part on the determined likelihood that the at least one of the target users will interact with the new digital content.