Emotion Analysis Using Multilayered Networks for Graphical Representation of Tweets

Anticipating audience reaction towards a certain piece of text is integral to several facets society ranging from politics, research, and commercial industries. Sentiment analysis (SA) useful natural language processing (NLP) technique that utilizes both lexical/statistical deep learning methods determine whether different sized texts exhibit positive, negative, or neutral emotion. However, there currently lack tools can be used analyze groups independent extract the primary emotion whole set. Therefore, current paper proposes novel algorithm referred as Multi-Layered Tweet Analyzer (MLTA) graphically models social media using multi-layered networks (MLNs) in order better encode relationships across sets tweets. Graph structures are capable capturing meaningful complex ecosystems compared other representation methods. State art Neural Networks (GNNs) information Tweet-MLN make predictions based on extracted graph features. Results show not only does MLTA predict larger set possible emotions, delivering more accurate sentiment standard negative neutral, it also allows for group-level Twitter data.