Emojis Segmentation from WhatsApp Chat Messages Using K-Means Clustering Technique for Students Sentiment Analysis

The ideas, attitudes, and thoughts expressed through social media networks are a major factor in sentiment analysis. Through WhatsApp chat messages; there has been a noticeable increase in the use of Emojis by students. Although the semantics and grammatical structure of sentiments based on simple Emojis communicate an extensive amount of information and are thought of as comprehensive communications. A single WhatsApp group message may include a variety of Emojis in addition to text. When used in a single message, numerous Emojis can express multiple emotions or sensations simultaneously, such as happiness, sadness, confusion, etc. In this situation, important areas or objects from message screenshots might be extracted and then subjected to sentiment analysis using image segmentation technique. This study employs a K-means clustering algorithm for segmentation of Emojis from the students chat messages. To assess the effectiveness of the system, ten frequently used Emojis are selected and experimented and obtained a total Emojis segmentation accuracy of 95.61%. The outcome of Emoji segmentation will be used for features extraction and classification of the students’ sentiments.