A Two-Level Learning Hierarchy of Nonnegative Matrix Factorization Based Topic Modeling for Main Topic Extraction

Abstract. Topic modeling is a type of statistical model that has been proven successful for tasks including discovering topics and their trends over time. In many applications, documents may be accompanied by metadata that is manually created by their authors to describe the se-mantic content of documents, e.g. titles and tags. A proper way of in-corporating this metadata to topic modeling should improve its perfor-mance. In this paper, we adapt a two-level learning hierarchy method for incorporating the metadata into nonnegative matrix factorization based topic modeling. Our experiments on extracting main topics show that the method improves the interpretability scores and also produces more in-terpretable topics than the baseline one-level learning hierarchy method