发明名称 Method and apparatus to identify outliers in social networks
摘要 A system that incorporates teachings of the present disclosure may include, for example, a process that obtains seed information including privacy profile settings of on-line social network users. The seed information is compared to a random population of on-line social network users and a sampling size of a total population is reduced according to the comparison. The reduced sampling is compared to one of a social graph, a generalized profile of an on-line user, or a combination thereof. The social graph is determined from a randomized-walk algorithm applied to an on-line social network of the users and an outlier is identified in the reduced sampling based on relationships between the on-line social network users. Additional embodiments are disclosed.
申请公布号 US9059897(B2) 申请公布日期 2015.06.16
申请号 US201414231964 申请日期 2014.04.01
申请人 AT&T INTELLECTUAL PROPERTY I, LP 发明人 Krishnamurthy Balachander
分类号 G06F15/173;H04L12/26;G06Q10/10 主分类号 G06F15/173
代理机构 Guntin & Gust, PLC 代理人 Guntin & Gust, PLC ;Trementozzi Ralph
主权项 1. A method, comprising: obtaining, by a system comprising a processor, seed information, wherein the seed information comprises privacy profile settings of on-line social network users; comparing, by the system, the seed information to a random population of on-line social network users, resulting in a comparison; reducing, by the system, a sampling size of a total population of on-line social network users according to the comparison, resulting in a reduced sampling of on-line social network users; comparing, by the system, the reduced sampling of on-line social network users to one of a social graph, a generalized profile of on-line social network users determined from the total population of on-line social network users, or a combination thereof; determining, by the system, the social graph from a randomized-walk algorithm applied to an on-line social network of the on-line social network users; and identifying, by the system, an outlier in the reduced sampling of on-line social network users based on relationships between the on-line social network users, and based on one of ambient or publicly available information combined with an intra-graph property of the on-line social network users, wherein the outlier does not conform to one of the social graph or the generalized profile of on-line social network users, wherein the randomized-walk algorithm comprises a Metropolis-Hastings algorithm, and wherein the seed information is received from equipment of a law enforcement agency.
地址 Atlanta GA US