TaxoKnow: Taxonomy as Prior Knowledge in the Loss Function of Multi-class Classification
Abstract
In this paper, we investigate the effectiveness of integrating a hierarchical taxonomy of labels as prior knowledge into the learning algorithm of a flat classifier. We introduce two methods to integrate the hierarchical taxonomy as an explicit regularizer into the loss function of learning algorithms. By reasoning on a hierarchical taxonomy, a neural network alleviates its output distributions over the classes, allowing conditioning on upper concepts for a minority class. We limit ourselves to the flat classification task and provide our experimental results on two industrial in-house datasets and two public benchmarks, RCV1 and Amazon product reviews. Our obtained results show the significant effect of a taxonomy in increasing the performance of a learner in semisupervised multi-class classification and the considerable results obtained in a fully supervised fashion.
Cite
@article{arxiv.2305.16341,
title = {TaxoKnow: Taxonomy as Prior Knowledge in the Loss Function of Multi-class Classification},
author = {Mohsen Pourvali and Yao Meng and Chen Sheng and Yangzhou Du},
journal= {arXiv preprint arXiv:2305.16341},
year = {2023}
}