Sentiment Analysis and Topic Modeling for Identifying Key Public Concerns of Water Quality/Issues

Abstract

A Slovakian proverb says, “Pure water is the World’s first and foremost medicine”. Public opinion on water quality can play a critical role in the policy process, primarily because water quality determines the quality of health. The goal of the paper is to analyze Twitter data to extract feelings and opinions into unstructured text. We perform the topic extraction that discovers the keywords in sentiments that capture the text’s recurring theme and is widely used to analyze large sets of sentiments to identify the most common topics quickly and efficiently. The main purpose of the article is to collect public opinion on Twitter to understand critical issues related to water quality. We used a few weeks of Twitter data and applied the principle of latent semantic analysis and decomposition of singular values to group key water quality questions that impact people’s lives. People do realise that bad water deteriorates the quality of life, particularly for children. Also, there is a looming fear about imminent water, which bothers them a lot. They are hopeful that some of the measures like the Swatch Bharat Mission, Rally for River, Water Kiosk and applying Machine Learning and Artificial Intelligence solutions to provide safe and clean water could solve the crisis.