公开(公告)号：US20140358673A1

公开(公告)日：2014-12-04

申请号：US13909315

申请日：2013-06-04

Applicant: Facebook, Inc.

Inventor： Richard Bill Sim ,  Jun Yang ,  Eitan Shay ,  Jiahui Wu ,  Spencer Powell ,  William Bullock

IPC: G06Q30/02

CPC classification number: G06Q30/0251 ,  G06Q30/0246 ,  G06Q30/0273 ,  G06Q50/01

Abstract: A group of social networking system users are associated with a holdout group for an advertisement. Users in the holdout group are not presented with the advertisement. When the advertisement is selected for presentation to a user, the social networking system presents the advertisement to the user if the user is not in the holdout group. However, if the user is in the holdout group, alternative content is presented to the user. If a user performs a conversion event associated with the advertisement via a client device, the social networking system determines a fee for an advertiser if the advertisement was presented to the user. The fee may be adjusted based on differences between conversion events by users in the holdout group for the advertisement and by users not in the holdout group.

Abstract translation: 一组社交网络系统用户与广告的保持组相关联。 保持组中的用户没有呈现广告。 当选择广告以呈现给用户时，如果用户不在保持组中，则社交网络系统向用户呈现广告。 然而，如果用户在保持组中，则替代内容被呈现给用户。 如果用户通过客户端设备执行与广告相关联的转换事件，则如果广告被呈​​现给用户，社交网络系统确定广告商的费用。 费用可以根据用于广告的保留组中的用户的转换事件和不在保留组中的用户之间的差异进行调整。

公开(公告)号：US20210168217A1

公开(公告)日：2021-06-03

申请号：US17173928

申请日：2021-02-11

Applicant: Facebook, Inc.

Inventor： William Bullock

IPC: H04L29/08 ,  G06N20/00 ,  G06F16/9537 ,  G06F16/29 ,  H04L29/12 ,  H04L12/58

Abstract: In one embodiment, a method includes generating predicted locations of each of a plurality of network addresses, wherein each predicted location is associated with a time stamp representing an age of the predicted location, determining a weighting factor representing a probability that at least one of the predicted locations of the network address corresponds to a true location of the network address based on location-related features associated with each network address and the time stamps, determining a weight for each predicted location based on at least the weighting factor, wherein the weight represents a probability that the predicted location corresponds to the true location of the network address, and providing one or more of the predicted locations that correspond to a particular network address based on the respective weights of the predicted locations in response to a request to identify a geographic location for the particular network address.

公开(公告)号：US10924560B2

公开(公告)日：2021-02-16

申请号：US16049712

申请日：2018-07-30

Applicant: Facebook, Inc.

Inventor： William Bullock

IPC: H04L29/08 ,  G06N20/00 ,  G06F16/9537 ,  G06F16/29 ,  H04L29/12 ,  H04L12/58 ,  G06F16/487 ,  H04W4/029

Abstract: In one embodiment, a method includes receiving one or more communication network addresses and one or more geographic locations of each network address, determining one or more location-related features based on each network address, generating one or more predicted locations of the network address, each predicted location corresponding to one of the first geographic locations of the network address, and each predicted location being associated with a time stamp representing an age of the predicted location, determining, based on the location-related features and the time stamps, a weighting factor representing a probability that at least one of the predicted locations of the network address corresponds to a true location of the network address, and determining, for each of the predicted locations, a weight based on at least the weighting factor, wherein the weight represents a probability that the predicted location corresponds to the true location of the network address.

公开(公告)号：US10277714B2

公开(公告)日：2019-04-30

申请号：US15592108

申请日：2017-05-10

Applicant: Facebook, Inc.

Inventor： William Bullock ,  Liang Xu ,  Li Zhou

IPC: G06F15/173 ,  H04L29/08 ,  G06K9/66 ,  G06F17/27 ,  G06F16/335 ,  G06Q30/02 ,  G06K9/00

Abstract: An online system predicts household features of a user, e.g., household size and demographic composition, based on image data of the user, e.g., profile photos, photos posted by the user and photos posted by other users socially connected with the user, and textual data in the user's profile that suggests relationships among individuals shown in the image data of the user. The online system applies one or more models trained using deep learning techniques to generate the predictions. For example, a trained image analysis model identifies each individual depicted in the photos of the user; a trained text analysis model derive household member relationship information from the user's profile data and tags associated with the photos. The online system uses the predictions to build more information about the user and his/her household in the online system, and provide improved and targeted content delivery to the user and the user's household.

公开(公告)号：US20180212931A1

公开(公告)日：2018-07-26

申请号：US15415684

申请日：2017-01-25

Applicant: Facebook, Inc.

Inventor： Li Zhou ,  William Bullock ,  Anh Phuong Bui

IPC: H04L29/06 ,  H04L29/08 ,  G06F21/62 ,  G06Q30/06

CPC classification number: H04L63/0428 ,  G06F21/6245 ,  G06Q30/0609 ,  G06Q30/0635 ,  H04L63/12 ,  H04L67/22 ,  H04L67/306

Abstract: Embodiments include one or more client devices accessible by users, an online system, and one or more partner systems such that the online system is able to identify a user of the online system across different devices and browsers based on the user activity that occurs external to the online system. A user performs user actions (e.g. purchase a product) on a web page of a partner system and may provide personally identifiable information (PII) to the partner system. The partner system provides the hashed PII and user actions performed by the user to the online system. The online system identifies a user profile on the online system by matching personal information in the user profile to the hashed PII. The online system generates a confidence score indicating a likelihood that the identified user of the online system is the individual that performed the external user action.

公开(公告)号：US20170068987A1

公开(公告)日：2017-03-09

申请号：US14847070

申请日：2015-09-08

Applicant: Facebook, Inc.

Inventor： Michael Daniel Levinson ,  William Bullock ,  Nathan John Davis ,  Chinmay Deepak Karande

IPC: G06Q30/02

CPC classification number: G06Q30/0246 ,  G06Q30/0275

Abstract: An advertisement system measures an ad lift metric for advertisement campaigns, which indicates the increase in conversions that can be attributed to the advertisement campaign. As impression opportunities become available for users for the ad in the lift study, the advertisement system determines whether the user is in a test group or a control group. To limit bias in the lift study, rather than holding out ads from being provided to users after the ad has been selected for the user and right before the impression, the system holds out the ads at a higher level in the ad selection process. In this manner, not all test group users receive the advertisement. The system computes the lift metric as e.g., the incremental lift (difference between conversion rates in the test and control groups), and this is divided by conversion rate of an exposed target group minus the incremental lift.

Abstract translation: 广告系统衡量广告活动的广告提升度量，这表明可以归因于广告活动的转化次数增加。 由于电梯研究中的广告的用户可以看到展示机会，所以广告系统确定用户是否在测试组或对照组中。 为了限制电梯研究中的偏见，而不是在为用户选择广告并在展示之前提供给用户，而是在广告选择过程中将广告提升到更高级别。 以这种方式，并非所有测试组用户都收到广告。 系统计算升力量度，例如增量提升（测试和控制组中转换率之间的差异），并将其除以暴露目标组的转化率减去增量提升。

公开(公告)号：US20160277526A1

公开(公告)日：2016-09-22

申请号：US14662124

申请日：2015-03-18

Applicant: Facebook, Inc.

Inventor： William Bullock ,  Carlos Gregorio Diuk Wasser

IPC: H04L29/08

CPC classification number: H04L67/306

Abstract: Systems, methods, and non-transitory computer-readable media can determine attribute information associated with attributes. The attribute information is associated with a first user and a second user. Match values for the attributes are determined based on the attribute information. A first rule is applied to the match values. The first user and the second user are predicted to be members in a first common household based on satisfaction of the first rule by the match values.

Abstract translation: 系统，方法和非暂时的计算机可读介质可以确定与属性相关联的属性信息。 属性信息与第一用户和第二用户相关联。 基于属性信息确定属性的匹配值。 第一条规则适用于匹配值。 基于匹配值对第一规则的满足，预测第一用户和第二用户是第一公共家庭中的成员。

公开(公告)号：US20140222583A1

公开(公告)日：2014-08-07

申请号：US13759976

申请日：2013-02-05

Applicant: Facebook, Inc.

Inventor： William Bullock ,  Sean Michael Bruich

IPC: G06Q30/02

CPC classification number: G06Q30/0269 ,  G06Q30/0241 ,  G06Q50/01

Abstract: An online system predicts values of a target characteristic for users in a set of users based on a reference set of users having known values for the target characteristic. Using descriptive characteristics of users in the reference set of users and target characteristic values for users in the reference set, the online system generates a model predicting values of the target characteristic based on user descriptive characteristics. The online system applies a global constraint on the target characteristic when generating the model, so the model extrapolates from the reference data while achieving aggregate results for values of the target characteristic that are consistent with the global constraint. The global constraint may be obtained from census data or another suitable global aggregate survey. Using the global constraint in the model avoids inaccuracies in reporting of user metrics.

Abstract translation: 在线系统基于具有目标特性的已知值的参考用户组来预测一组用户中的用户的目标特征的值。 使用参考用户组中的用户的描述特征和参考集中的用户的目标特征值，在线系统基于用户描述特征生成预测目标特征值的模型。 在生成模型时，在线系统对目标特征应用全局约束，因此模型将从参考数据中推算出来，同时获得与全局约束一致的目标特征值的聚合结果。 全球约束可以从普查数据或另一个合适的全球总体调查中获得。 在模型中使用全局约束避免了用户指标报告中的不准确。