An image analysis approach to text analytics based on complex networks

Text network analysis has received increasing attention as a consequence of its wide range of applications. In this study, we extend a previous work founded on the study of topological features of mesoscopic networks. Here, the geometrical properties of visualized networks are quantified by using several image analysis techniques. Such properties are used to probe the networks characteristics in terms of authorship. It was found that the visual features account for performance similar to that achieved by using topological measurements. Also, we combined and compared the two types of features, topological and geometrical, and the results suggest that the information provided by network topology and image features are complementary.