DEEP LINGUISTIC AND BEHAVIORAL ANALYSIS FOR FAKE PROFILE IDENTIFICATION IN ONLINE SOCIAL NETWORKS

Abstract

The increasing presence of fake profiles in online social networks poses serious challenges to user trust, information authenticity, and digital security. These profiles exploit human and algorithmic vulnerabilities to spread misinformation, perform scams, and manipulate public opinion. This research presents a deep linguistic and behavioral analysis framework for the identification of fake profiles using a combination of Machine Learning (ML) and Natural Language Processing (NLP) techniques. The proposed system extracts textual, emotional, and structural features from user data to capture both linguistic inconsistencies and abnormal behavioral patterns. Supervised ML algorithms are trained on large-scale datasets to classify accounts as genuine or deceptive. Experimental results reveal that the fusion of language-based and behavioral attributes enhances accuracy and generalization across diverse social media platforms. The proposed approach demonstrates the potential of intelligent hybrid systems to strengthen the integrity and safety of online communities.