Injurious Comment Detection and Removal utilizing Neural Network

There are a lot of ways to communicate in this cyber world. With this increasingly growing era there is also much obstruction in a safe and secure environment. There has been an exponential growth in cyber bullying and abusing. Deep learning methods have recently begun to be used to detect abusive comments made in online forums. Detecting, and classifying, online abusive language is a non-trivial NLP challenge because online comments are made in a wide variety of contexts, and contain words from many different formal and informal lexicons. For this to overcome we design a model that detects the level of toxicity in a message and replaces it with another phrase. It uses a Deep Neural network model that takes a message/comment as an input and checks for various parameters such as Toxic, Severe Toxic, Identity hate, threat, etc. And the application finally then replaces the portion with another word/phrase. Examining things, you care about can be troublesome. The danger of misuse and provocation online implies that numerous individuals quit communicating and offer up on looking for changed thoughts. Stages battle to adequately encourage discussions, driving numerous networks to restrict or totally shut down client remarks.