\(\textbf{E}^{3}\) -MG: End-to-End Expert Linking via Multi-Granularity Representation Learning

Expert linking is a task to link any mentions with their corresponding expert in a knowledge base (KB). Previous works that focused on explicit features did not fully exploit the fine-grained linkage and pivotal attribute inside of each expert work, which creates a serious semantic bias. Also, such models are more sensitive to specific experts resulting from the isolationism for class-imbalance instances. To address this issue, we propose \(\mathbf {E^{3}}\) -MG (End-to-End Expert Linking via Multi-Granularity Representation Learning), a unified multi-granularity learning framework, we adopt a cross-attention module perceptively mining fine-grained linkage to highlight the expression of masterpieces or pivotal support information and a multi-objective learning process that integrates contrastive learning and knowledge distillation method is designed to optimize coherence between experts via document-level coherence. E \(^3\) -MG enhances the representation capability of diverse characteristics of experts and demonstrates good generalizability. We evaluate E \(^3\) -MG on KB and extern datasets, and our method outperforms existing methods.