A Survey on Entity Linking Based on Deep Neural Networks
Entity Linking(EL)aims to link mentions in text with their corresponding entities in knowledge base.In recent years,the rapid development of deep learning imporves the performance of entity linking.This paper system-atically reviews recent entity linking models and methods from the four aspects:① General framework including candidate entity generation,candidate entity ranking and unlinkable mention prediction.②Low-resource entity link-ing including cross-lingual and cross-domain transfer methods.③Domain specific entity linking with a focus on bio-medical and social media domains.④ Multimodal entity linking is described.Finally,the technical challenges faced by entity linking are analyzed and the future development is prospected.
万方数据
维普
