Automatic classification of text documents has recently been a hot topic in research. Information retrieval, machine learning, and natural language processing (NLP) techniques are required for proper classification of text materials. Our goal is to concentrate on three major approaches to automatic text classification using machine learning techniques: supervised, unsupervised, and semi-supervised. This paper reviews various text categorization algorithms using the machine learning paradigm in this study. We hope that our research will shed light on the relationships between various text classification techniques as well as the future research trend in this field.