Fake accounts can be created by humans, computers, or even cyborgs. A cyborg account is halfhuman and half-bots. A person physically initiates the account, but a bot handles all subsequent activities. There are differences between bots versus human profiles. When profiles are fake and not hacked by authorized users,bots are referred to as Sybil profiles. The research is recommended using a Twitter dataset, which consolidates real and Fake profile and implements that dataset using various methods like machine learning, natural language processing, and deep learning. Using these datasets, examination of distinct exploratory analyses to distinguish semantic properties broadly present in unreliable content. Furthermore correlate our Fake profile identification standards precision with expected accuracy to put our consequences in aspect. Utilization of Natural Language Processing, machine learning, and deep learning procedures to complete our models and distinguish which models will give higher efficiency. The CNN has a training accuracy of 99 percent accuracy

Fake profile, feature selection, classification, deep learning, fake user, ICA, and CNN.