Files
Abstract
A well-known fact of the internet age is that online social media is accessed regularly by an increasing number of users. Such platforms enable its users to create, share and spread information with ease in real time. In this research, we explore the possibility of harnessing this information to identify incidents of harmful algal blooms on water bodies. We target the information shared by the users of a popular micro-blogging service known as Twitter. We propose a way to annotate the slew of information obtained from these platforms to create ground truth and for analysis. We develop and test a platform that can extract and separate tweets that report incidents of harmful algal blooms. We apply Machine Learning and Natural Language Processing techniques to identify locations of such reported incidents if mentioned in the body of the tweet. An exploratory quantitative analysis of the collected data is also presented.