Comparison of technologies for data exploration and network analysis with volcano data

Abstract

The goal of the preliminary network analysis is to find and confirm the known results from the exploratory steps which in this case is showing how many volcanoes are in each country. Network analysis is applied to additional dimensions of the data to show complementing graphs exploring the relationship between volcanoes and their dominant rock types. Network analysis allows for the derivation of new insights which are more difficult to portray with the tabular formats in the form of network graphs. This is done by creating a second network to include dominant rock types. A third network is created showing the general distribution of volcanoes by country colored by the dominant rock types to show multiple-variables being displayed in a network analysis. Data visualization using Tableau and R complement each other for use with geophysical datasets and research in applying network concepts for volcano visualization can be used for data analysis. Applying network concepts to volcano visualization can be for data exploration and data analysis and be used across other geophysical datasets.

The purpose of this work is to compare the use of technologies for data exploration and data analysis within the field of volcanology. In both exploration and formal analysis, volcanic visualizations apply novel data analysis techniques to geological data. The research presented herein contains exploratory analysis of data on numerous attributes of volcanoes, world-wide from the Smithsonian Institution Museum of Natural History Global Volcanism Program. This analysis includes a tabular exploration and visualizations of volcano data using Tableau. This data exploration provides research familiarity with the dataset using tabular methods. Network analysis provides quantitative and qualitative insights for the complex, high-dimensional data relationship based on attributes not usually considered in low-dimensional studies [9]. After cleaning the data to modify the names of duplicate volcanoes to be unique, network analysis is performed on the data.