Suhani V Prajapati

Predicting Drug Abuse Beahviour using Deep Learning

Twitter data is collected using the twitter APIs, also our dataset is targeted towards the students from renowned universities.Thereafter preprocessing of the tweets is done in order to obtain a clean data.Word embedding is done using BERT to extract the features. The unlabeled tweets would be fed into the traditional binary classifier in order to classify it into two possible classifications. We would be using Naive Bayes Classification algorithm.Later, these tweets would be used to analyse the sentiments of scholars there after predict the mental health.