Provides a workflow for classifying scientific literature abstracts using positive-unlabeled (PU) learning. Supports abstract ingestion, text preprocessing, vocabulary construction, document-term matrix creation, and model fitting via the PLUS algorithm.