Performing Text Classification and Categorization through Unsupervised Learning

Text classification is a very crucial step in applying Natural Language Processing (NLP) to fathom real-life situations in networking domains such as engineering, business, and science. It is proven to be outright practical and potent in inscribing problems with unlabeled data. Today, there are undeniable unsupervised learning algorithms that are requisite in allocating document texts such as hierarchical clustering, k-means clustering, and probabilistic clustering. This paper illustrates and proves that the text in each online document is classified and categorized into several different sets through unsupervised techniques. Furthermore, this research determines the highest accuracy among the unsupervised methods. In addition, the research identifies and resolves possible problems that the unsupervised learning models encounter from text classification and categorization.