Abstract: A method trains a model for providing content items to users of a social networking system. The system generates profile vectors based on user profile information such as demographic data and personal data. The system logs actions performed by users on the social networking system and generates behavior vectors based on the logged actions. The profile vectors and behavior vectors are each associated with a user attribute, e.g., the age or gender of a user. The system generates a difference vector based on a profile vector and a behavior vector. The difference vector is then used as a feature to train the model using machine learning techniques. The trained model may select content items that a target user is most likely to be interested in and interact with.

Abstract: An online system automatically and dynamically determines an audience for content by clustering users across various dimensions, and refining targeting criteria for the content. The online system receives content and initial targeting criteria from a content provider. The content is provided to a group of users that meet the initial targeting criteria. The system collects content response data from the group of users that were provided the content, including user responses to the content and dynamic data relating to time and location of the user responses. The content response data is further integrated with user characteristics, content presentation data, and social response data to generate integrated user-content data of the content. Clusters of users are generated based on features of the integrated user-content data, and refined targeting criteria are identified based on the generated clusters that can then be used for more accurate targeting of the content to users.

Abstract: A method trains a model for providing content items to users of a social networking system. The system generates profile vectors based on user profile information such as demographic data and personal data. The system logs actions performed by users on the social networking system and generates behavior vectors based on the logged actions. The profile vectors and behavior vectors are each associated with a user attribute, e.g., the age or gender of a user. The system generates a difference vector based on a profile vector and a behavior vector. The difference vector is then used as a feature to train the model using machine learning techniques. The trained model may select content items that a target user is most likely to be interested in and interact with.