基于非平衡学习与交互式标注的引文情感动机标注系统

Citation Emotive Tagging System based on Unbalanced Learning and Interactive Tagging

随着现代文献的数据量不断增长,仅仅依靠论文引用次数的传统方式无法很好的描述论文的真实影响力,基于引文内容的情感动机的研究成为新的研究重点。虽然当前的研究中存在大量自然语言处理工具可以处理文献文本,但并没有专门针对文献引文情感动机的处理工具。为解决引文动机样本分布极度不均衡和引文动机标注数据生成效率低下的问题,本文提出构建一个基于非平衡学习和交互式标注的引文系统。经过实验证明,本文提出的系统可以较好提高引文情感动机标注效率。

With the increasing data volume of modern literature,the traditional method of paper citation cannot fully represent the real influence of the paper,so emotional motivation(Emotive)study based on citation content has become a new research focus.Although a large number of natural language processing tools exist in the current research to process literature texts,special tools to process literature citation emotional motivation have not been found yet.In order to solve the problems of extremely unbalanced sample distribution of citation motivation and low efficiency of citation motivation annotation data generation,this paper proposes building a citation system based on unbalanced learning and interactive annotation.Experiment results show that the system proposed in this paper can improve the tagging efficiency of citation emotive.