Tag cloud visualization in a pie chart manner.
By Stefan Jänicke.
Widely used and perceived as being fun, tag clouds are an important component in the social web to visualize summaries of textual data in an intuitive, lucid manner. Popular techniques like Wordle compute aesthetic layouts for a set of tags representing a specific data facet, but only few approaches are capable of visualizing multiple data facets.
However, these approaches were not suitable for various research questions arisen from the digital humanities project eXChange, which explores the meaning and usage of specific terms in ancient texts. Using a keyword based search when working with digital corpora, humanities scholars often receive numerous results in form of text passages, which are hard to revise individually, and a generation of valueable hypotheses is nearly impossible. Therefore, we were asked for a visualization that gives an overview of the results and can be used to directly jump to text passages containing specific co-occurrences of the searched term. Moreover, the scholars wished to compare the results of multiple queries within one interface in order to be able to discover similarities and differences of contexts in which related terms were used.
By introducing TagPies, we present a tag cloud design that supports the humanities scholars in examining their research questions and generating valueable hypotheses. The TagPies JavaScript library — based upon the d3-cloud — generates tag clouds in a pie chart manner capable of visualizing and comparing various data facets.
TagPies is licensed under the latest Version of the Fair Academic License (FAL).
FAL for TagPies: http://www.tagpies.vizcovery.org/source.html
More information about FAL can be found under http://www.vizcovery.org/fal.html.