摘要：Personification is important to ensure safety in social media. Images embedded in social media data is an important source intelligently ensure the personal safety. Paper proposes an applied machine learning on embedded image media to extract important feature traits and understand the semantics of correlation of feature space to understand the personification of given user. A strong belief is calculated using cascaded method and this inference is applied to different ml algorithms to derive the overlap of feature space. This algorithm validates the proposed objective that the user with more overlap and correlation among feature space are highly personified with each other. Using this personification-based method, better safety is ensured to mitigate possible social media attacks.
关键词：Persona Identification;social media;Embedded Media
