Neural Cross-Lingual Coreference Resolution and its Application to\n Entity Linking

We propose an entity-centric neural cross-lingual coreference model that\nbuilds on multi-lingual embeddings and language-independent features. We\nperform both intrinsic and extrinsic evaluations of our model. In the intrinsic\nevaluation, we show that our model, when trained on English and tested on\nChinese and Spanish, achieves competitive results to the models trained\ndirectly on Chinese and Spanish respectively. In the extrinsic evaluation, we\nshow that our English model helps achieve superior entity linking accuracy on\nChinese and Spanish test sets than the top 2015 TAC system without using any\nannotated data from Chinese or Spanish.\n