On the Influence of Hyperparameters in Tree-Based Linear Methods for Extreme Multi-Label Text Classification: Insights for Efficient and Effective Search
This study innovatively analyzes the hyperparameters of tree-based linear methods and suggests an efficient and effective guideline that leads to consistent improvements across datasets, thereby strengthening tree-based linear methods as a stronger XMTC baseline.
Extreme multi-label classification (XML) and hierarchical text classification (HTC) address closely related multi-label prediction problems, but have largely developed as separate research areas. XML focuses on very large label spaces and typically evaluates ranked label lists, while HTC assumes a human-curated label h...
Florian Hauss, Tom Speier, Nerijus Bertalis et al.· Proceedings of the 2026 ACM...· 0 citations
The rapid growth of user-generated textual content on the internet has intensified the need for accurate and scalable text classification methods. However, supervised learning approaches remain heavily constrained by the high cost and effort required for manual data annotation, particularly in large and heterogeneous d...
Large-scale multi-label text classification assigns a small subset of relevant labels to each document from a vocabulary containing thousands or tens of thousands of candidate labels. Although pretrained language models have improved semantic text representations, most representation-based approaches center their predi...
Hui Ye, Jing Zhang, Xiu-Long Yang et al.· 0 citations
Multi-label classification (MLC) requires predicting multiple relevant labels for each instance, where a central challenge is modeling complex label dependencies arising from co-occurrence patterns. Existing approaches are limited in capturing high-order label correlations, relying on implicit learning through contrast...
Pei-Yu Zhang, Heng Ping, Nikos Kanakaris et al.· 0 citations
Data-Informed Centroid Splitting (DICS), a clustering-based framework that constructs a compact and informative set of candidate splits using data-driven priors, significantly reduces the split search space for classification tasks while preserving predictive performance.
