The Social Net of Sentiment: Improving the Base Sentiment Analysis of High Impact Events with Lexical Category Exploration

Social networking has created a large source of peoples' opinions and statements available on the internet. This is particularly true in the case of high impact events, such as the Parkland school shooting in Florida. This paper approaches the analysis of high impact events in two ways. First, tweet sentiment analysis using the NLTK machine learning standard with TextBlob is applied. This approach is then augmented with lexical categorical analysis using the Python tool Empath as an added analysis step. The TextBlob standards are compared to Empath's sentiment analysis results to compare the accuracy of the two methods. The paper presents this combined approach to improve sentiment analysis by using Empath as an added analysis step and briefly discuss future further refinements.