A Framework for Youth Sentiment Analysis Using Natural Language Processing

Abstract

Social networks are currently the most widely used platforms for exchanging thoughts on various subjects or events, particularly those geared at young people. Consequently, the Natural Language Processing (NLP) industry has access to an abundant source of data provided by those applications. This paper presents a model for sentiment analysis using Naive Bayes algorithm. A proposed model is applied to two dataset samples to train, create, and test classification models. The supervised approach combined unigram for feature extraction and the Naive Bayes algorithm to extract the trending topics for youth Tweets. The model is evaluated on a test set during the worldwide crisis and is shown to be effective in predicting opinions on new reviews. The results of the evaluation demonstrate that the model is capable of accurately predicting opinions with a high degree of accuracy.