Techniques for probabilistic identification for content delivery

A computing device comprising:

• a processing circuit;
• and
• a memory having executable instructions stored thereon, which when executed by the processing circuit cause the processing circuit to:
• receive a request from a web server for content;
• determine, based on the request, a set of user-probability pairs, wherein each of the user-probability pairs includes a predicted user identifier and a probability that the predicted user identifier is associated with the request from the web server, and wherein each pair in the set of user-probability pairs includes an attribution value associated therewith;
• apply an objective function to the set of user-probability pairs to select one of the user identifiers with a corresponding candidate user profile, wherein applying the objective function includes the processing circuit being caused to:
• in a first selection, select a random user identifier and candidate user profile;
• or
• in a second selection, select the user identifier and candidate user profile having a greatest attribution value associated therewith;
• wherein the processing circuit is further caused to make a decision to execute the first selection or the second selection based on a selection parameter;
• access the candidate user profile of the selected user identifier to determine a set of attributes of the candidate user profile;
• select, based on the request and the set of attributes of the candidate user profile of the selected user identifier, a content item from a plurality of content item options to be sent to the web server;
• and
• send the content item to the web server.

A non-transitory computer-readable storage medium having executable instructions stored thereon, which when executed by a processing circuit of a computing device cause the computing device to:

• receive a request from a web server for content;
• determine, based on the request, a set of user-probability pairs, wherein each of the user-probability pairs includes a predicted user identifier and a probability that the predicted user identifier is associated with the request from the web server, and wherein each pair in the set of user-probability pairs includes an attribution value associated therewith;
• apply an objective function to the set of user-probability pairs to select one of the user identifiers with a corresponding candidate user profile, wherein applying the objective function includes the computing device being caused to:
• in a first selection, select a random user identifier and candidate user profile;
• or
• in a second selection, select the user identifier and candidate user profile having a greatest attribution value associated therewith;
• wherein the computing device is further caused to make a decision to execute the first selection or the second selection based on a selection parameter;
• access the candidate user profile of the selected user identifier to determine a set of attributes of the candidate user profile;
• select, based on the request and the set of attributes of the candidate user profile of the selected user identifier, a content item from a plurality of content item options to be sent to the web server;
• and
• send the content item to the web server.