公开(公告)号：US20190197400A1

公开(公告)日：2019-06-27

申请号：US15855950

申请日：2017-12-27

Applicant: Facebook, Inc.

Inventor： Ying Zhang ,  Wenhai Yang ,  Ou Jin

IPC: G06N3/08 ,  G06F17/30

CPC classification number: G06N3/08 ,  G06F16/951 ,  G06Q50/01

Abstract: In one embodiment, a method includes accessing an input vector representing an input post, wherein the input post includes one or more n-grams and an image, the input vector corresponds to a point in a d-dimensional vector space, the input vector was generated by an artificial neural network (ANN) based on a text vector representing the one or more n-grams of the input post and an image vector representing the image of the input post; and the ANN was jointly trained to receive a text vector representing one or more n-grams of a post and an image vector representing an image of the post and then output a probability that the received post is related to the training posts of a training page; and determining a topic of the input post based on the input vector.

公开(公告)号：US20190197190A1

公开(公告)日：2019-06-27

申请号：US15855942

申请日：2017-12-27

Applicant: Facebook, Inc.

Inventor： Ying Zhang ,  Wenhai Yang ,  Ou Jin

IPC: G06F17/30 ,  G06N3/08

CPC classification number: G06F16/9535 ,  G06F16/24575 ,  G06F16/24578 ,  G06F16/248 ,  G06N3/08

Abstract: In one embodiment, a method includes accessing a user profile associated with a user of an online social network, wherein the user profile identifies one or more topics that the user is interested in; accessing post vectors, wherein each post vector represents one of a plurality of posts, indicates one or more topics, and for each of the topics, indicates a probability that the post is related to the corresponding topic; ranking the posts based on comparisons between the user profile and the post vectors; and providing for display to the user posts based on the ranking.

公开(公告)号：US10002329B2

公开(公告)日：2018-06-19

申请号：US14498898

申请日：2014-09-26

Applicant: Facebook, Inc.

Inventor： Hussein Mohamed Hassan Mehanna ,  Stuart Michael Bowers ,  Alexandre Defossez ,  Parv Oberoi ,  Ou Jin

IPC: G06N99/00

CPC classification number: G06N20/00

Abstract: An online system simplifies modification of features used by machine learned models used by the online system, such as machined learned models with high dimensionality. The online system obtains a superset of features including features used by at least one machine learned model and may include additional features. From the superset of features, the online system generates various groups of features for a machine learned model. The groups of features may be a group including features currently used by the machine learned model, a group including all available features, and one or more intermediate groups. Intermediate groups include various numbers of features from the set selected based on measures of feature impact on the machine learned model associated with various features. A user may select a group of features, test the machine learning model using the selected group, and then launch the tested model based on the results.

公开(公告)号：US10943171B2

公开(公告)日：2021-03-09

申请号：US15694742

申请日：2017-09-01

Applicant: Facebook, Inc.

Inventor： Qiang Wu ,  Ou Jin ,  Liang Xiong

IPC: G06N3/04 ,  G06N3/08 ,  G06T1/40 ,  G06T1/20 ,  G06T1/60 ,  G06F9/00

Abstract: An optimized computer architecture for training an neural network includes a system having multiple GPUs. The neural network may be divided into separate portions, and a different portion is assigned to each of the multiple GPUs. Within each GPU, its portion is further divided across multiple training worker threads in multiple processing cores, and each processing core has lock-free access to a local parameter memory. The local parameter memory of each GPU is separately, and individually, synchronized with a remote master parameter memory by lock memory access. Each GPU has a separate set of communication worker threads dedicated to data transfer between the GPU and the remote parameter memory so that the GPU's training worker threads are not involved with cross GPU communications.

公开(公告)号：US20200045354A1

公开(公告)日：2020-02-06

申请号：US16054871

申请日：2018-08-03

Applicant: Facebook, Inc.

Inventor： Tianshi Gao ,  Xiangyu Wang ,  Ou Jin ,  Yifei Huang ,  Vignesh Ramanathan

IPC: H04N21/25 ,  G06N3/04 ,  G06N3/08 ,  H04N21/24 ,  H04N21/234 ,  G06K9/46

Abstract: An online system receives content items from a third party content provider. For each content item, the online system inputs an image into a neural network and extracts a feature vector from a hidden layer of the neural network. The online system compresses each feature vector by assigning a label to each feature value representing whether the feature value was above a threshold value. The online system identifies a set of content items that the user has interacted with and determines a user feature vector by aggregating feature vectors of the set of content items. For a new set of content items, the online system compares the compressed feature vectors of the content item with the user feature vector. The online system selects one or more of the new content items based on the comparison and sends the selected content items to the user.

公开(公告)号：US20190197399A1

公开(公告)日：2019-06-27

申请号：US15855946

申请日：2017-12-27

Applicant: Facebook, Inc.

Inventor： Ying Zhang ,  Wenhai Yang ,  Ou Jin

IPC: G06N3/08 ,  G06F17/30

CPC classification number: G06N3/08 ,  G06F16/951 ,  G06Q50/01

Abstract: In one embodiment, a method includes accessing an input vector representing an input post, wherein: the vector space comprises clusters each associated with a topic; each cluster was determined based on a clustering of training-page vectors corresponding to training pages that each comprise training posts, each training post submitted by a user to a training page and comprises content selected by the user; and each training-page vector was generated by an ANN that was trained, based on the training posts of training pages associated with the ANN, to receive a post and then output a probability that the received post is related to the training posts of the training pages; determining that the input vector is located within a particular cluster in the vector space; and determining a topic of the input post based on the topic associated with the particular cluster that the input vector is located within.

公开(公告)号：US20150332317A1

公开(公告)日：2015-11-19

申请号：US14279149

申请日：2014-05-15

Applicant: Facebook, Inc.

Inventor： Xinyi Cui ,  Wenjie Fu ,  Haomin Yu ,  Ou Jin ,  Eitan Shay ,  Richard Bill Sim ,  Jun Yang

IPC: G06Q30/02 ,  G06Q50/00

CPC classification number: G06Q30/0275 ,  G06Q30/0246 ,  G06Q30/0247 ,  G06Q50/01

Abstract: An advertising system receives from an advertiser at a social networking system an advertisement request, the advertisement request comprising advertisement content and a specification of a target audience for the advertisement content. The advertising system defines a plurality of bid values for the advertisement request. For each of the plurality of bid values, the advertisement system estimates a corresponding value of advertisement reach for the target audience, for example, by estimating a number of users of the target audience for each of whom the given bid value is expected to have resulted in at least one successful impression. The advertiser is provided a visual representation of a bid-reach landscape representing the estimated plurality of advertisement reach values as a function of the plurality of bid values. The advertising system provides, to the advertiser, one or more recommendations for bid values for which corresponding return-on-investment metrics exceed a specified threshold.

Abstract translation: 广告系统在社交网络系统中从广告商接收广告请求，广告请求包括广告内容和广告内容的目标受众的规范。 广告系统定义广告请求的多个出价值。 对于多个投标值中的每一个，广告系统估计目标受众的广告到达的相应值，例如，通过估计预期给出的投标价值的每个的目标受众的用户数量 至少有一个成功的印象。 向广告商提供表示作为多个投标值的函数的估计的多个广告到达值的出价达标景观的视觉表示。 广告系统向广告商提供一个或多个关于投标价值的建议，其中相应的投资回报率度量超过指定的阈值。

公开(公告)号：US10229357B2

公开(公告)日：2019-03-12

申请号：US14851336

申请日：2015-09-11

Applicant: Facebook, Inc.

Inventor： Ou Jin ,  Stuart Michael Bowers ,  Dmytro Dzhulgakov

IPC: G06N3/08 ,  G06N99/00

Abstract: The present disclosure is directed to a high-capacity training and prediction machine learning platform that can support high-capacity parameter models (e.g., with 10 billion weights). The platform implements a generic feature transformation layer for joint updating and a distributed training framework utilizing shard servers to increase training speed for the high-capacity model size. The models generated by the platform can be utilized in conjunction with existing dense baseline models to predict compatibilities between different groupings of objects (e.g., a group of two objects, three objects, etc.).

公开(公告)号：US20170076198A1

公开(公告)日：2017-03-16

申请号：US14851336

申请日：2015-09-11

Applicant: Facebook, Inc.

Inventor： Ou Jin ,  Stuart Michael Bowers ,  Dmytro Dzhulgakov

IPC: G06N3/08 ,  G06N99/00

CPC classification number: G06N99/005

Abstract: The present disclosure is directed to a high-capacity training and prediction machine learning platform that can support high-capacity parameter models (e.g., with 10 billion weights). The platform implements a generic feature transformation layer for joint updating and a distributed training framework utilizing shard servers to increase training speed for the high-capacity model size. The models generated by the platform can be utilized in conjunction with existing dense baseline models to predict compatibilities between different groupings of objects (e.g., a group of two objects, three objects, etc.).

Abstract translation: 本公开涉及可以支持大容量参数模型（例如，具有100亿个权重）的高容量训练和预测机器学习平台。 该平台实现了联合更新的通用特征转换层和利用碎片服务器的分布式培训框架，以提高高容量模型大小的培训速度。 由平台生成的模型可以与现有的密集基线模型一起使用，以预测不同对象组之间的兼容性（例如，一组两个对象，三个对象等）。

公开(公告)号：US20150332310A1

公开(公告)日：2015-11-19

申请号：US14279131

申请日：2014-05-15

Applicant: Facebook, Inc.

Inventor： Xinyi Cui ,  Wenjie Fu ,  Haomin Yu ,  Ou Jin ,  Eitan Shay ,  Richard Bill Sim ,  Jun Yang

IPC: G06Q30/02 ,  G06Q50/00

CPC classification number: G06Q30/0275 ,  G06Q30/0206 ,  G06Q30/0242 ,  G06Q30/0244 ,  G06Q30/0246 ,  G06Q30/0283 ,  G06Q50/01

Abstract: An advertising system predicts advertisement reach for a received advertisement request based on an advertiser-specified bid amount and a specification of a target audience. The system samples the target audience, and for each sampled user of the target audience, accesses a recent impression history to obtain costs or bids associated with recent advertisement impressions. The system compares the advertiser-specified bid amount in the received advertisement request to costs or bid values associated with successful advertisement impressions, for each sampled user, in order to determine whether the received advertisement request would have won a bid auction for each given sampled user to successfully reach each given sampled user. An estimated aggregate reach for the sampled users is computed and extrapolated to the targeted user population to estimate a total reach of the advertisement content for the target audience.

Abstract translation: 广告系统基于广告商指定的出价金额和目标受众的规格来预测所接收的广告请求的广告到达。 系统对目标受众进行采样，对于目标受众的每个抽样用户，访问最近的展示历史记录以获取与最近广告展示相关联的成本或出价。 系统将收到的广告请求中的广告商指定的投标金额与成功的广告印象相关联的成本或出价值进行比较，以便确定所接收到的广告请求是否已经为每个给定的抽样用户赢得了竞标拍卖 成功到达每个给定的采样用户。 计算样本用户的估计总收视率并将其推广到目标用户群体，以估计目标受众的广告内容的总覆盖面。