Spherical Double K-Means: a co-clustering approach for text data analysis

In this study, we introduce the Spherical Double K-Means (SDKM) clustering method for text data. A novel approach for simultaneous clustering of terms and documents. Using the strengths of k-means, double k-means, and spherical k-means, SDKM addresses the challenges of high dimensionality, noise, and sparsity inherent in text analysis. We address the choice of the number of clusters, both for the words and documents, using the cluster validity index pseudo-F, and verify the reliability of the method through simulation studies. We apply SDKM to the corpus of US presidential inaugural addresses, spanning from George Washington in 1789 to Joe Biden in 2021. Our analysis reveals distinct clusters of words and documents that correspond to significant historical themes and periods, showcasing the method's ability to facilitate a deeper understanding of the data. Our findings demonstrate the efficacy of SDKM in uncovering underlying patterns in textual data.