Open AccessPersona Identification based on Social Media Profile Images for Personification and Safety
Author(s) -
Prema Pandurang Gawade,
Sarang Achyut Joshi
Publication year - 2022
Publication title -
webology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.259
H-Index - 18
ISSN - 1735-188X
DOI - 10.14704/web/v19i1/web19087
Subject(s) - social media , feature (linguistics) , persona , computer science , identification (biology) , inference , space (punctuation) , correlation , artificial intelligence , semantics (computer science) , feature vector , computer vision , pattern recognition (psychology) , human–computer interaction , world wide web , mathematics , philosophy , linguistics , botany , geometry , biology , programming language , operating system
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.