Machine Learning and Sentiment Analysis: Examining the Contextual Polarity of Public Sentiment on Malaria Disease in Social Networks

摘要

Malaria, a major deadly disease which is still a threat to human life's even though numerous efforts has been put to fight it, still affects over two hundred million people each year amongst which over a million individuals dies. Twitter happens to be an important and comprehensive source of information that is quite subjective to individual sentiments towards public health care. In this study, we extracted tweets from the social network twitter, we preprocessed the tweets extracted and built a model to fit our data using a machine learning approach for text classification to determine the contextual polarity of every tweet on the subject of malaria in the bid to harvest peoples' opinion towards malaria and understand how well research and recent development in the aid to tackle malaria has affected the opinions of the public towards the subject malaria. This study finds that tweets extracted, preprocessed and classified in this study were majorly classified as negative (-ve) due to the fact that tweets tweeted were majorly about different occurrence of death, misinformation and need for donations to save a life, hence a major awareness is needed.