Hierarchical Text Classification: a review of current research

t is often the case that collections of documents are annotated with hierarchically-structured concepts. However, the benefits of this structure are rarely taken into account by commonly-used classification techniques. Conversely, Hierarchical Text Classification methods are devisedto take advantage of the labels’ organization to boost classification performance. With this work,we aim to deliver an updated overview of current research in this domain. We begin by definingthe task and framing it within the broader text classification area, examining important shared concepts such as text representation. Then, we dive into details regarding the specific task,providing a high-level description of its traditional approaches. We then summarize recentlyproposed methods, highlighting their main contributions. We additionally provide statisticsfor the most adopted datasets and describe the benefits of using evaluation metrics tailored to hierarchical settings. Finally, a selection of recent proposals is benchmarked against non-hierarchical baselines on five domain-specific datasets.