Abstract:
Analysis of Social Media Platforms provides helpful information for users on social media.
Recent papers about user interaction on social media explore methods for predicting user
interaction. These analyses of Social Media Platforms have included Active Users and
Youth Users analysis based on year. Yet, the studies have not incorporated text data. This
research explores the usefulness of incorporating text data to predict user interaction. The
study incorporates two types of machine learning models: Linear Regression, Time Series
Model(Arima). The models are unique in their use of the data. The research collects 208
Users based on social media platforms such as Facebook, Instagram, YouTube, LinkedIn,
Twitter, Netflix, Whatsapp, Snapchat, Google+, TikTok, Pinterest etc per year. The models
learn and test on Youth users in order to predict user interaction. The study found that text
data produced the best models. The research further demonstrates that Time series models
perform best.
