Title:Topological Data Analysis on Simple English Wikipedia Articles

Abstract:Single-parameter persistent homology, a key tool in topological data analysis, has been widely applied to data problems, with statistical techniques that quantify the significance of the results. In contrast, statistical techniques for two-parameter persistence, while highly desirable for real-world applications, have scarcely been considered. We present three statistical approaches for comparing geometric data using two-parameter persistent homology, and we demonstrate the applicability of these approaches on high-dimensional point-cloud data obtained from Simple English Wikipedia articles. These approaches rely on the Hilbert function, matching distance, and barcodes obtained from two-parameter persistence modules computed from the point-cloud data. We demonstrate the applicability of our methods by distinguishing certain subsets of the Wikipedia data, and by comparison with random data. Results include insights into the construction of null distributions and stability of our methods with respect to noisy data. Our statistical methods are broadly applicable for analysis of geometric data indexed by a real-valued parameter.