Description

Mallet is a Java-based statistical natural language processing package designed for application developers. It enables users to gather and analyze the collections of unlabeled text, handle distinct tasks such as tokenizing strings, removing stopwords, and converting sequences into count vectors, perform sampling-based implementations of Latent Dirichlet Allocation, Pachinko Allocation, and Hierarchical LDA, transform text documents into numerical representations, divide documents into tokens using a regular expression, as well as distinguish between uses of words with multiple meanings. Features include document classification, topic modeling, information extraction, numerical optimization, machine learning, and sequence tagging.