The Application of Ensemble Learning on Named Entity Recognition for Legal Knowledgebase of Properties Involved in Criminal Cases
Zongshen Jiang · 2020
Entities are important units that carry information in texts and the core units of the knowledgebase. Named entity recognition (NER) plays a key role in the construction of knowledgebase. In this paper we explore the NER problem for legal knowledgebase of properties involved in criminal case. In the scenario with limited training corpus and high recognition performance requirements, parallel ensemble learning is introduced on the basis of four NER methods. Experimental results show that ensemble learning can further improve the accuracy of recognition.