IDENTIFICATION OF TOXIC TWEET

Abstract

The objective ofthis project is to develop an algorithm that can be used to identify the hateful comment from the lot. This report spots out how identify these types oftweets and also what points toxic or we can normally looking at when discussing such cases. Different stages involving text processing like the pre-processing stage, segmentation and feature extraction will be studied and discussed. Finally, product of the algorithms will be written. This projegrtises data form the twitter and then analysis it for toxic or simply bad comments that may harm anyone or fulfils the intention of harming anyone. The project are we the end Neural Network model to classify the basic word encodings including Universal sentence encodings TensorFlow libraries to develop such system. The system first proceeds gathering ofdata and requirement ofdata involves a tweeter developers account and a working scraping code for tweeter. Filtering, segmentation, resizing and features extraction are also performed in the process. Thisls the process of creating this classifier, and recommendations for future development and conclusions included in the report.