Affinity-driven blog cascade analysis and prediction

• Autores: Hui Li, Sourav Bhowmick, Aixin Sun, Jiangtao Cui
• Localización: Data mining and knowledge discovery, ISSN 1384-5810, Vol. 28, Nº 2, 2014, págs. 442-474
• Idioma: inglés
• Texto completo no disponible (Saber más ...)
• Resumen

  • Information propagation within the blogosphere is of much importance in implementing policies, marketing research, launching new products, and other applications. In this paper, we take a microscopic view of the information propagation pattern in blogosphere by investigating blog cascade affinity. A blog cascadeis a group of posts linked together discussing about the same topic, and cascade affinityrefers to the phenomenon of a blog’s inclination to join a specific cascade. We identify and analyze an array of macroscopicand microscopiccontent-oblivious features that may affect a blogger’s cascade joining behavior and utilize these features to predict cascade affinity of blogs. Based on these features, we present two non-probabilistic and probabilistic strategies, namely support vector machine (SVM) classification-based approach and Bipartite Markov Random Field-based (BiMRF) approach, respectively, to predict the probability of blogs’ affinity to a cascade and rank them accordingly. Evaluated on a real dataset consisting of 873,496 posts, our experimental results demonstrate that our prediction strategy can generate high quality results ($$F1$$-measure of 72.5 % for SVM and 71.1 % for BiMRF) comparing with the approaches using traditional or singular features only such as elapsed time, number of participants which is around 11.2 and 8.9 %, respectively. Our experiments also showed that among all features identified, the number of quasi-friendsis the most important factor affecting bloggers’ inclination to join cascades.