Modeling the structure and evolution of discussion cascades

TL;DR

This paper models the structure and growth of online discussion threads across four websites using a biased preferential attachment model, revealing common statistical patterns and communication behaviors.

Contribution

It introduces a novel maximum likelihood estimation method for preferential attachment models, effectively capturing the evolution of discussion cascades across diverse platforms.

Findings

The model accurately reproduces degree and subtree size distributions.

Discussion cascades exhibit similar statistical properties despite platform differences.

Parameters reveal insights into user communication habits.

Abstract

We analyze the structure and evolution of discussion cascades in four popular websites: Slashdot, Barrapunto, Meneame and Wikipedia. Despite the big heterogeneities between these sites, a preferential attachment (PA) model with bias to the root can capture the temporal evolution of the observed trees and many of their statistical properties, namely, probability distributions of the branching factors (degrees), subtree sizes and certain correlations. The parameters of the model are learned efficiently using a novel maximum likelihood estimation scheme for PA and provide a figurative interpretation about the communication habits and the resulting discussion cascades on the four different websites.