Analyzing textual data by multiple word clouds

Abstract

Word clouds are a good way to represent textual data with meta information. However, when it comes to analyzing multiple data sources, they are difficult to use. This is due to their poor comparability. The proposed RadCloud merges multiple standard clouds into a single one, while retaining the information about the origin of the words. Using separate clouds as well as this new visualization technique, a tool for textual data analysis is created: MuWoC. The resulting software is then evaluated by using it with multiple data sources ranging from encyclopedia articles, over literature to custom CSV files.