Abstract: Techniques for forecasting for an advertisement campaign are described. A personalized communication system can receive a request for an advertisement campaign on a social network. The request can have a member attribute and a time frame. The personalized communication system can access member data and behavior data from the social network. Additionally, the personalized communication system can determine a target group based on the member data and the member attribute. Furthermore, the personalized communication system can calculate a number of unique visitors to the social network from the target group based on the member attribute, the time frame, and a frequency cap. Subsequently, the personalized communication system can forecast a number of messages for the first advertisement campaign based on the calculated number of unique visitors, the behavior data, and the time frame.

Abstract: A method of optimizing online advertising campaign allocations is disclosed. It is determined that an auction for a set of advertising slots has been triggered. It is identified that the advertising campaigns are configured to bid on the set of advertising slots. A ranking score for each of the advertising campaigns is determined. The ranking scores are adjusted for each cost-per-click advertising campaign of the set of advertising campaigns by an adjustment factor specific to a context of the auction. The set of advertising slots is allocated to the winners of the auction. The winners of the auction are communicated for integration into a content page.

Abstract: The disclosed embodiments provide a method and system for managing online advertising. During operation, the system obtains a constraint on a first metric associated with an online advertising auction. Next, the system uses a computer to generate a tradeoff curve of the first metric and a second metric associated with the online advertising auction and use the tradeoff curve to obtain a value for the second metric that meets the constraint on the first metric. The system then uses the value to manage the online advertising auction.

Abstract: Systems and methods for placing ads in a block on a webpage are disclosed. Generally, two ranking models are trained using a first and second ads data set. The first model predicts a first click probability for each ad in the first ads data and rank the ads based on the eCPM. The second model is trained using the second ads data set comprising a subset of the first ads data set and interaction features related to ad position in the block. The second model predicts a second click probability for each ad in the second ads data set. An overall expected revenue for each arrangement of ads in the second ads data set is then calculated. The computer system selects the arrangement with maximum computed overall expected revenue and places the ads in the block on the webpage according to the selected arrangement.

Abstract: This disclosure relates to systems and methods that include obtaining from a member activity database, with a processor, data indicative of past interactions by members of a social network with content items associated with the social network, the content items including sponsored content items and organic content items. A positioning condition for content items is determined for individual members, based on the data as obtained, including sponsored content items and organic content items, on a user interface displayed to the individual members that is estimated to meet a optimization criterion, the optimization criterion including engagement with the content items by members individually and by members overall. The user interface is provided to one of the members based, at least in part, on the positioning condition.

Abstract: Systems and methods for placing ads in a block on a webpage are disclosed. Generally, two ranking models are trained using a first and second ads data set. The first model predicts a first click probability for each ad in the first ads data and rank the ads based on the eCPM. The second model is trained using the second ads data set comprising a subset of the first ads data set and interaction features related to ad position in the block. The second model predicts a second click probability for each ad in the second ads data set. An overall expected revenue for each arrangement of ads in the second ads data set is then calculated. The computer system selects the arrangement with maximum computed overall expected revenue and places the ads in the block on the webpage according to the selected arrangement.

Abstract: Systems and methods for placing ads in a block on a webpage are disclosed. Generally, two ranking models are trained using a first and second ads data set. The first model predicts a first click probability for each ad in the first ads data and rank the ads based on the eCPM. The second model is trained using the second ads data set comprising a subset of the first ads data set and interaction features related to ad position in the block. The second model predicts a second click probability for each ad in the second ads data set. An overall expected revenue for each arrangement of ads in the second ads data set is then calculated. The computer system selects the arrangement with maximum computed overall expected revenue and places the ads in the block on the webpage according to the selected arrangement.

Abstract: Systems and methods are described that select advertisements for placement on a series of consecutively-accessed web pages, such as consecutively-accessed web search results pages generated in response to a particular search. The systems and methods perform a separate advertisement ranking process to select advertisements for placement on each web page in the series of consecutively-accessed web pages at the time the web page is accessed. For web pages that follow the first web page in the series, the systems and methods utilize an advertisement ranking technique that calculates a probability that a user will select an advertisement based on certain user selection feedback features. The user selection feedback features for an advertisement are determined by comparing attributes of the advertisement to attributes of user-selectable items that were presented on one or more of the previously-accessed web pages in the series that are known to have been selected or not selected by the user.

Abstract: A system and method is described herein that selects advertisements for two or more regions of a web page based on rankings of the advertisements generated using different models. By ranking advertisements separately for each region of a web page, and choosing highest ranked advertisements for each region, advertisements can be selectively chosen such that a user is more likely to select an advertisement in each region of the web page. As a result, the user experience can be enhanced and the advertising revenue can be correspondingly increased.

Abstract: A system and method is described herein that selects advertisements for two or more regions of a web page based on rankings of the advertisements generated using different models. By ranking advertisements separately for each region of a web page, and choosing highest ranked advertisements for each region, advertisements can be selectively chosen such that a user is more likely to select an advertisement in each region of the web page. As a result, the user experience can be enhanced and the advertising revenue can be correspondingly increased.

Abstract: Systems and methods are described that select advertisements for placement on a series of consecutively-accessed web pages, such as consecutively-accessed web search results pages generated in response to a particular search. The systems and methods perform a separate advertisement ranking process to select advertisements for placement on each web page in the series of consecutively-accessed web pages at the time the web page is accessed. For web pages that follow the first web page in the series, the systems and methods utilize an advertisement ranking technique that calculates a probability that a user will select an advertisement based on certain user selection feedback features. The user selection feedback features for an advertisement are determined by comparing attributes of the advertisement to attributes of user-selectable items that were presented on one or more of the previously-accessed web pages in the series that are known to have been selected or not selected by the user.

Abstract: Systems and methods for placing ads in a block on a webpage are disclosed. Generally, two ranking models are trained using a first and second ads data set. The first model predicts a first click probability for each ad in the first ads data and rank the ads based on the eCPM. The second model is trained using the second ads data set comprising a subset of the first ads data set and interaction features related to ad position in the block. The second model predicts a second click probability for each ad in the second ads data set. An overall expected revenue for each arrangement of ads in the second ads data set is then calculated. The computer system selects the arrangement with maximum computed overall expected revenue and places the ads in the block on the webpage according to the selected arrangement.