Machine Learning-Based Prediction of Users' Involvement on Social Media

Abstract

Useful information can be extracted through the analysis of Facebook posts. Text analysis and image analysis can play a vital role towards this. To predict the users' involvement, text data and image data can be incorporated using some machine learning models. These models can be used to perform testing on advertisements that are posted on Facebook for users' involvement prediction. Count of share and comments with sentiment analysis are included as users' involvement. This chapter contributes to understand the users' involvement on social media along with finding out the best machine learning model for prediction of users' involvement. The procedure of prediction with both text data and image data by suitable models is also discussed. This chapter produces a predictive model for posts of Facebook to predict users' involvement that will be based on the number of shares and comments on the post. The best models are obtained by using the combination of image data and text data. Further, it demonstrated that random models are surpassed by the models that are integrated for prediction.