A review of hierarchical classification methods for text documents with a comparative perspective

Nowadays, a vast amount of information and knowledge exists in textual form and digital format. Given the continuous growth of knowledge, document classification has become crucial for reducing information complexity and enabling easy and rapid access. In real-world scenarios, many text document classification problems exhibit a hierarchical nature. Flat classification algorithms are often inadequate for such tasks, both because they ignore relationships between categories and due to the large number of categories involved. To address these challenges, hierarchical classification methods are employed. This article reviews distinct approaches for hierarchical classification of text documents and presents a comparative analysis of their advantages and disadvantages.