Abstract: Users of a social networking system are assigned to households using prediction models that rely, in part, on user profile information and social graph data. Information about users may be received by a social networking system through various channels (e.g., declared/profile information, user history, IP addresses, Global Positioning System (GPS) data from check-in events and/or continuously provided by mobile devices, external household information, and/or social information). Scoring models may use statistical analysis of the received user information to predict household membership for users. User attributes, such as previous names, date of birth, social graph data, locations, life events, and check-ins, may be factors in generating confidence scores of predicted household memberships. Weighted scoring models may use machine learning methods for measuring the accuracy of the household membership prediction.

Abstract: Users of a social networking system are assigned to households using prediction models that rely, in part, on user profile information and social graph data. Information about users may be received by a social networking system through various channels (e.g., declared/profile information, user history, IP addresses, Global Positioning System (GPS) data from check-in events and/or continuously provided by mobile devices, external household information, and/or social information). Scoring models may use statistical analysis of the received user information to predict household membership for users. User attributes, such as previous names, date of birth, social graph data, locations, life events, and check-ins, may be factors in generating confidence scores of predicted household memberships. Weighted scoring models may use machine learning methods for measuring the accuracy of the household membership prediction.

Abstract: An advertisement impression management system receives an audience list that identifies a set of users associated with an interest topic. The system also identifies whether users associated with advertisements displayed by advertising publishers are in the received audience list. Advertising impression data associated with the displayed advertisements is received by the system to identify users accessing the displayed advertisements. The system identifies unique users based on online tracking methods and user information extracted from user databases of a social networking system. Correction techniques and external panel data are used to correct user information of the identified unique users to improve accuracy. An audience report is generated by the system to evaluate the ability and effectiveness of the advertising publishers in reaching members of the users in the audience list.

Abstract: Users of a social networking system are assigned to households using prediction models that rely, in part, on user profile information and social graph data. Information about users may be received by a social networking system through various channels (e.g., declared/profile information, user history, IP addresses, Global Positioning System (GPS) data from check-in events and/or continuously provided by mobile devices, external household information, and/or social information). Scoring models may use statistical analysis of the received user information to predict household membership for users. User attributes, such as previous names, date of birth, social graph data, locations, life events, and check-ins, may be factors in generating confidence scores of predicted household memberships. Weighted scoring models may use machine learning methods for measuring the accuracy of the household membership prediction.

Abstract: A social networking system or other user registration site builds a log of exposures by users to advertisements outside of the user registration site to determine their effectiveness. For each user exposed to an advertisement, a log entry is created indicating that the user has been exposed to the advertisement. Tracking pixels are embedded into advertisements that, when accessed, enable the social networking system or user registration site to log access to the advertisement by the user. From the log files, the user registration site identifies exposed users and selects unexposed users with similar demographics and/or behavior information to generate a control group. The two groups can be surveyed about the advertisement to determine its effectiveness. The user exposure information also can be used to retarget advertisements, to measure advertising effectiveness on connections of exposed users, and to measure actions of exposed users.

Abstract: In one embodiment, a social networking system models a number of exposures to an advertisement for a concept for a set of users, sample from the set of users attitudinal data toward the concept, and determine effectiveness of the advertisement by evaluating the attitudinal data against the number of exposures to the advertisement.

Abstract: Embodiments of the invention combine information from different data sets, such as social networks, advertising networks, and/or panels, each data set comprising statistics about past viewership of content (e.g., advertisements). The result of the combination is a model that, when applied to statistics about viewing of particular content, produces viewing statistics about the particular content that are more accurate than the data of any given one of the different data sets when taken in isolation.

Abstract: Actions and/or behaviors of social networking system users are observed and used for measuring advertising effectiveness. More specifically, advertisements from an advertising campaign are selectively targeted and presented to specific subsets of social network users and withheld from other subsets of social network users. After the advertisements are presented, actions performed by users in the different subsets are be identified and analyzed to determine metrics describing the effectiveness of the particular advertising campaign.

Abstract: Embodiments of the invention combine information from different data sets, such as social networks, advertising networks, and/or panels, each data set comprising statistics about past viewership of content (e.g., advertisements). The result of the combination is a model that, when applied to statistics about viewing of particular content, produces viewing statistics about the particular content that are more accurate than the data of any given one of the different data sets when taken in isolation.

Abstract: An online advertising system receives ads and ad exposure goals, such as a desired number of impressions or amount of presentation time, from advertisers. The online advertising system may also receive a time-based advertising purchase. A request for an ad is received for a client, and an amount of exposure is determined for each received ad across a plurality of ad mediums. The ad mediums may include both controlled ad mediums and external ad mediums. The advertising system selects an ad for presentation among the received ads based on the received ad exposure goals associated with the ads and the determined amount of exposure for each ad. The advertising system may select an ad by determining bids for each ad and conducting an auction among the ads. In such an embodiment, the advertising system may determine the bids for ads that have met their ad exposure goals to be zero.

Abstract: An online advertising system receives ads and ad exposure goals, such as a desired number of impressions or amount of presentation time, from advertisers. The online advertising system may also receive a time-based advertising purchase. A request for an ad is received for a client, and an amount of exposure is determined for each received ad across a plurality of ad mediums. The ad mediums may include both controlled ad mediums and external ad mediums. The advertising system selects an ad for presentation among the received ads based on the received ad exposure goals associated with the ads and the determined amount of exposure for each ad. The advertising system may select an ad by determining bids for each ad and conducting an auction among the ads. In such an embodiment, the advertising system may determine the bids for ads that have met their ad exposure goals to be zero.

Abstract: Embodiments of the invention combine information from different data sets, such as social networks, vendor systems, and/or panels, each data set comprising statistics about past consumer behavior (e.g., product purchases). The result of the combination is a model that, when applied to statistics about purchases of a particular product, produces predicted consumer behavior statistics about the particular product that are more accurate than the data of any given one of the different data sets when taken in isolation.

Abstract: In one embodiment, a social networking system models a number of exposures to an advertisement for a concept for a set of users, sample from the set of users attitudinal data toward the concept, and determine effectiveness of the advertisement by evaluating the attitudinal data against the number of exposures to the advertisement.

Abstract: Embodiments of the invention combine information from different data sets, such as social networks, advertising networks, and/or panels, each data set comprising statistics about past viewership of content (e.g., advertisements). The result of the combination is a model that, when applied to statistics about viewing of particular content, produces viewing statistics about the particular content that are more accurate than the data of any given one of the different data sets when taken in isolation.

Abstract: A social networking system or other user registration site builds a log of exposures by users to advertisements outside of the user registration site to determine their effectiveness. For each user exposed to an advertisement, a log entry is created indicating that the user has been exposed to the advertisement. Tracking pixels are embedded into advertisements that, when accessed, enable the social networking system or user registration site to log access to the advertisement by the user. From the log files, the user registration site identifies exposed users and selects unexposed users with similar demographics and/or behavior information to generate a control group. The two groups can be surveyed about the advertisement to determine its effectiveness. The user exposure information also can be used to retarget advertisements, to measure advertising effectiveness on connections of exposed users, and to measure actions of exposed users.