Human-Machine Collaboration Based Named Entity Recognition

Named Entity Recognition (NER) is an important task in Natural Language Processing (NLP), and its main goal is to extract the required entities from a given text and label the entity type. Not only is an important research topic, but its identification quality and efficiency also have an important impact on follow-up tasks (such as machine translation, intelligent question answering system, etc.). Current research methods are primarily based on Deep Learning (DL) models. In practical applications, such models often rely on a large number of labeled data, which will show a certain limitation in applications in specific domains. At the same time, due to the high uncertainty of human language and the openness involving problems, the DL model does not reach the point of complete replacement of humanity. Human-machine collaboration refers to that in the changing environment, human and machines perform tasks alternately, so that the task can achieve the best performance with a small amount of human participation. In this paper, we propose a Human-Machine Collaboration based Named Entity Recognition (HMCNER) model by utilizing the complex cognitive reasoning ability of human beings and combining with existing DL model. The extensive experimental results show that our model can efficiently complete the NER task based on the existing research results, and have practical application.