Exodus Bi-Directional LSTM Based Multivariate Texmoji Classification

Stories that contain text and emojis manifests humans' emotions in a very broad way. Emojis mixed with regular words on many social media sites has changed how we talk. It lets people send feelings without having to describe them in a standard way. The proposed work here mainly focuses to collect such kind of texts combined with the emojis and predict the 8 basic human emotions revealed through it i.e., Anger, fear, sadness, disgust, surprise, anticipation, trust and joy. The collected instances are summarized with transformers summarization algorithm, which then we calculate the emotional intensity of each emotion and perform a multi-label classification using Bi-Directional LSTM network, and rank the emotions as per the predicted emotional inteneity.