Title:	TWITTER TEMPORAL EVOLUTION ANALYSIS: COMPARING EVENT AND TOPIC DRIVEN RETWEET GRAPHS
Author(s):	Giambattista Amati, Simone Angelini, Francesca Capri, Giorgio Gambosi, Gianluca Rossi, Paola Vocca
ISBN:	978-989-8533-54-8
Editors:	Piet Kommers and Guo Chao Peng
Year:	2016
Edition:	Single
Keywords:	Graph analysis, social media, Twitter graph, retweet graph, graph dynamics
Type:	Full Paper
First Page:	155
Last Page:	162
Language:	English
Cover:
Full Contents:	click to dowload
Paper Abstract:	We study the retweet graphs evolution over time. We compare two different retweet graphs: the event driven retweet graph, filtered by topics about specific events (i.e. the Black Friday 2015 and the World Series 2015) and the sampling retweet graph, filtered by language (i.e. Italian) from the whole Twitter stream. To obtain the Italian Twitter sample we use a list of the most used Italian stop words and the Twitter native selection function for languages. We analyze the evolution of these retweet graphs over a period of two months, and compare the main structural measures that are generally used to characterize the nature of graphs: average distance, clustering coefficient, max in-out degree, max in-degree and out-degree, the size of the largest component, number of connected components. Results show a significant difference between these two type of graphs, both in the way they grow, and in the way the above measures evolve. In particular, the sampling retweet graphs dimensions (edges and vertices sizes) and the considered measures grow almost linearly in the observation period, while the event driven graphs show a skewed distribution and reach a predictably saturation point at end of the event. The measures for the sampling retweet graph are more similar to a social network, such as Facebook, than to an information network, whilst the event driven retweet graphs is the opposite. The study of the temporal evolution and the classification of different Twitter graphs is a preliminary work in order to better understand the nature of Twitter, how trends evolve over time, to detect both authoritative and spamming accounts, and to derive a suitable mathematical evolutionary model of Twitter communities.
	Go Back