Exploring Tabular Data Through Networks

Abstract

Representing and visualizing data as networks is a widely spread approach to analyzing highly connected data in domains such as medicine, social sciences, and information retrieval. Investigating data as networks requires pre-processing, retrieval or filtering, conversion of data into networks, and application of various network analysis approaches. These processes are usually complex and hard to perform without some programming knowledge and resources. To the best of our knowledge, most solutions attempting to make these functionalities accessible to users focus on particular processes in isolation without exploring how these processes could be further abstracted or combined in a real-world application to assist users in their data exploration and knowledge extraction journey. Furthermore, most applications focusing on such approaches tend to be closed-source. This paper introduces a solution that combines the approaches above as part of Collaboration Spotting X (CSX), an open-source network-based visual analytics tool for retrieving, modeling, and exploring or analyzing data as networks. It abstracts the concepts above through the use of multiple interactive visualizations. In addition to being an easily accessible open-source platform for data exploration and analysis, CSX can also serve as a real-world evaluation platform for researchers in related computer science areas who wish to test their solutions and approaches to machine learning, visualizations, interactions, and more in a real-world system.