Input dataset has metadata and textual data in German language.
Overview- To generate embeddings, BERT technique and its variants are used, Clustering the samples based on each variant of traits into clusters and finding the correlation between each of the clusters.