论文部分内容阅读
[目的 /意义]对引文文本分类的标准、实现方法和应用进行梳理,分析存在的问题,提出可改进的方向。[方法/过程]总结目前引文文本分类的几个重要角度,如基于引用功能、基于情感倾向、基于引文影响力等,对引文文本分类的实现方法进行比较,分析其优缺点。[结果 /结论]目前引文文本没有统一的分类标准和实现方法,引文文本的获取较为困难,计算机分类算法准确率较低,中文引文文本分析文献少。未来研究思路和方向应该是:统一文本分类的标准,提高引文文本计算机处理技术的准确性,扩大应用范围。
[Purpose / Significance] To sort out the criteria, implementation methods and applications of the citation text classification, analyze the existing problems and propose the direction for improvement. [Methods / Processes] To summarize several important perspectives of current citation text categorization, such as based on quoting function, based on affective tendencies, based on citation influence and so on, compare the implementation methods of citation text categorization and analyze its advantages and disadvantages. [Results / Conclusions] There is no uniform classification standard and implementation method in current citation texts. It is difficult to obtain citation texts. Computer classification algorithms have low accuracy and few citation texts in Chinese citation. Future research ideas and directions should be: the standard unified text classification to improve the accuracy of citation text computer processing technology to expand the scope of application.