Easy Abstract

We present BoostXML, a deep learning-based XML method enhanced greatly by gradient boosting. In BoostXML, we pay more attention to tail labels in each boosting step by optimizing the residual mostly from unfitted training instances with tail labels. A corrective step is further proposed to enable the cooperation between the text encoder and weak learners, which reduces the risk of falling into local optima and improves model performance.

Get the paper

• The paper will be provided if published.