A Multi-Criteria Decision Analysis Approach for Predicting User Popularity on Social Media

Social media platforms like Facebook and Twitter have enabled users to connect and share information on an unprecedented scale. However, the extensive adoption of these platforms has led to only a select few gaining mass popularity. Identification of popular social media influencers is crucial for applications like marketing and recommendations. But manual assessment of user popularity is challenging due to numerous multivariate indicators of influence. In this paper, we propose an intelligent system to predict social media user popularity by applying Multi-Criteria Decision Analysis (MCDA). Our approach selects key popularity criteria based on profile data from Facebook and Twitter like follower count, friends, tweets etc. We construct a comparison matrix that systematically assigns weights to each criterion according to its importance for determining popularity. The weighted criteria are then utilized in a prediction algorithm that collects user data via APIs, normalizes it and computes an overall popularity score. We discuss the system design and key technical details of the platform. The structured weighting technique and flexible data-driven prediction algorithm offer an automated and scalable means to identify top social media influencers. We highlight the ability to customize the criteria comparisons and algorithm for different contexts. The proposed approach advances capabilities for analyzing social media user influence.