Improved Accuracy for Exploring Text - Based Emotion Recognition in Social Media Conversation Generalized Linear Model Compared with Decision Tree

AIM: The objective of this work is text-based emotion recognition of social media conversation by comparing Novel Generalized linear model (GLM) and Decision Tree Algorithm. Materials and Methods: The Generalized linear model sample size=10 and Decision Tree sample size=10 the estimation has been done iteratively to obtain the accuracy of 98.01% Results: Generalized linear model(GLM) methodology produced better accuracy of 93.01% when correlated with the Decision Tree accuracy algorithm. In this experiment, p=0.828 (p<0.05 ) it is statistically insignificant parameter with a pretest power of 80% is used. Conclusion: For exploring text - based emotion recognition in social media conversation the Generalized Linear Model (GLM) performed significantly with better accuracy than Decision Tree algorithm.