Classification of Emojis using Artificial Neural Network and Natural Language Processing
Projit Chandra, Upasana Prasad · International Conference on Computing for Sustainable Global Development · 2021
This paper presents a generalized system which can be integrated with any machine learning or deep learning classification algorithm to find the intended interpretation of an emoji in a sentence. The primary purpose of this system is to detect and translate the offensive or the sexual meanings represented by the emojis into English words so that it can be classified by any classifier. The system is designed to store and retrieve the emojis and their meanings into a hierarchical lookup data structure. This data structure uses natural language processing to translate an emoji or a string of emojis into the meaning that best fits the sentence in which they were in. As a result, this system was able to determine the possible meaning from all the plausible meaning of an emoji or a string of emojis with respect to the statement. These results were further supported and validated by the high accuracy of the deep learning classifier used to classify the translated sentences.