Augmented machine learning ensemble extension model for social media health trends predictions

Abstract

Social Networks are the source of rich, interactive, textual, and other media.Users of the social media generate data at a tremendous pace.This data consisting of user opinions and attitudes is so large that it has necessitated automated methods to analyze and extract knowledge from the same. Social networks have been studied and analyzed using various graph-based analysis techniques. Prominent analysishas centered on features like ego-networks, distance, centrality, sub-networks etc. The areas of study for social media analysis have been centered around populations, boundaries, Cohesion, Centrality and Brokerage, Prestige and Ranking. In the past several models have been propounded for various machine learning based analytics for the Social Networks study but there is a perceived need for studying social networks for health data using Ensemble Learning wherein an array of various Machine Learning techniques can be employed to achieve better classification or clustering results. We introduce an Analytical Model which will identify most discussed terms/ topics of health/ healthcare on social networks to predict the emerging health trends. The model is to use temporal datasets to deduce multi-label classification of health-related topics. The Model employs the technique of Temporal Clustering (using Machine Learning) on the Topic Classification done on datasets using Ensemble Machine Learning to deduce the most discussed topics. Using this model, we will see how Ensemble Machine Learning based Analytical Model for analyzing social network data for health topics is efficient than traditional Machine Learning technique(s).