Pinpoint Influential Posts and Authors

TL;DR

This paper introduces an analytical model that uses topic modeling and TF-IDF to identify influential social media posts and authors, aiding in monitoring impactful online discussions and customer relationship management.

Contribution

The paper presents a novel combination of Latent Dirichlet Allocation and TF-IDF for pinpointing influential posts and authors in social web analysis.

Findings

Effective identification of influential posts in social media

Model successfully monitors evolving online 'storms' impacting audiences

Potential for extending analysis to broader data sources

Abstract

This research presents an analytical model that aims to pin-point influential posts across a social web comprised of a corpus of posts. The model employs the Latent Dirichlet Al-location algorithm to associate posts with topics, and the TF-IDF metric to identify the key posts associated with each top-ic. The model was demonstrated in the domain of customer relationship by enabling careful monitoring of evolving "storms" created by individuals which tend to impact large audiences (either positively or negatively). Future research should be engaged in order to extend the scope of the corpus by including additional relevant publicly available sources.