Near Real Time Twitter Sentiment Analysis and Visualization

Abstract

Twitter can be considered as a large scale network. People's opinions matter a lot to analyze how knowledge spreads impact lives. In this project, we took advantage of the Apache Spark Streaming fast and memory computing platform to retrieve live tweets and perform sentiment analysis. The primary purpose is to provide a tool to evaluate the score of sentiments in streams. This paper reports on the nature of an analysis of emotions, collecting vast numbers of tweets. Results identify the view of users through tweets into positive, neutral and negative about coronavirus. This project on Spark Streaming to analyze tweets, hashtags or specific keyword/keywords such as (corona) from live twitter data streams. Data is collected from input sources like Twitter and processed downstream using Spark Streaming. Then, how sentiment scores can be generated for tweets and build visualization dashboards on the data using Elasticsearch and Kibana.