PREDICTING GEN-Z PERSONALITY ON TWITTER BASED ON BIG FIVE MODEL WITH KNN AND SVM

Abstract

Generation Z is a group that is very connected to digital technology, especially social media such as Twitter. Their widespread presence on these platforms creates a unique opportunity to understand their behavioural patterns and personalities. However, research on personality prediction on social media is still limited and focused on certain platforms or different age groups. Personality prediction can help to find out someone's personality by just looking at tweets on social media. This research aims at two things: first, to build a Gen-Z personality prediction model on Twitter based on the Big Five Personality Model with the K-Nearest Neighbor (KNN) algorithm and Support Vector Machine (SVM). Second, test and compare the performance of previously generated personality prediction models with various evaluation metrics. The research results show that the KNN algorithm has an accuracy rate of 0.73%, precision of 0.73%, recall of 0.73%, and score of 0.72%. Based on the test results, the SVM algorithm obtained the best accuracy, which received an accuracy of 0.78%, precision of 0.82%, recall of 0.78%, and F1-score of 0.78%. This research contributes in two ways: first, scientifically, by understanding Gen-Z personalities on Twitter, and second, by developing new prediction methods and insights into Gen-Z behaviour. Second, practically, by helping with communication and marketing strategies, product/service development and social interventions for Gen-Z.

Keywords

• machine learning
• prediction algorithm
• big five personality
• knn
• svm