公开(公告)号：US20150332336A1

公开(公告)日：2015-11-19

申请号：US14290355

申请日：2014-05-29

Applicant: Facebook, Inc.

Inventor： Sue Ann Hong ,  Gang Wu ,  Rituraj Kirti ,  Vladislav Isenbaev ,  Boris Pierre Arnoux

IPC: G06Q30/02 ,  G06Q50/00 ,  G06F17/30

Abstract: A social networking system selects advertisements for a user based on user characteristics of the user in response to a request to present an advertisement to the user. To increase the number of advertisements eligible for presentation to the user, the social networking system associates the user with one or more cluster groups associated with targeting criteria that are not satisfied by the user's characteristics. To determine whether to associate a user with a cluster group, the social networking system determines a cluster score for the cluster group based on the user's characteristics. If the cluster score equals or exceeds a cluster cutoff score for the cluster group, the user is associated with the cluster group. The cluster cutoff score may be determined based on an estimated distribution of users so that a target number or percentage of users have cluster scores less than the cluster cutoff score.

Abstract translation: 响应于向用户呈现广告的请求，社交网络系统基于用户的用户特征为用户选择广告。 为了增加有资格呈现给用户的广告的数量，社交网络系统将用户与与用户的特征不满足的定位标准相关联的一个或多个集群组相关联。 要确定是否将用户与群集组关联，社交网络系统将根据用户的特征确定群集组的群集分数。 如果集群分数等于或超过集群组的集群截断分数，则用户与集群组关联。 可以基于用户的估计分布来确定群集截断分数，使得目标数量或用户百分比具有小于群集截断分数的聚类分数。

公开(公告)号：US10509816B2

公开(公告)日：2019-12-17

申请号：US14290355

申请日：2014-05-29

Applicant: Facebook, Inc.

Inventor： Sue Ann Hong ,  Gang Wu ,  Rituraj Kirti ,  Vladislav Isenbaev ,  Boris Pierre Arnoux

IPC: G06Q30/02 ,  G06Q50/00 ,  G06F17/30 ,  G06F16/435 ,  G06F16/28 ,  G06F16/2457

Abstract: A social networking system selects advertisements for a user based on user characteristics of the user in response to a request to present an advertisement to the user. To increase the number of advertisements eligible for presentation to the user, the social networking system associates the user with one or more cluster groups associated with targeting criteria that are not satisfied by the user's characteristics. To determine whether to associate a user with a cluster group, the social networking system determines a cluster score for the cluster group based on the user's characteristics. If the cluster score equals or exceeds a cluster cutoff score for the cluster group, the user is associated with the cluster group. The cluster cutoff score may be determined based on an estimated distribution of users so that a target number or percentage of users have cluster scores less than the cluster cutoff score.