机构地区: 南开大学商学院信息资源管理系,天津300071
出 处: 《情报理论与实践》 2017年第9期105-111,76,共8页
摘 要: [目的/意义]创新是学术论文的本质要求,如何有效评价学术论文的创新力一直以来备受国内外专家和学者关注。随着信息技术的发展,利用计算机技术从论文内容角度对单篇学术论文的创新力进行评价逐步成为可能。[方法/过程]文章提出了一种基于研究水平的单篇学术论文创新力评价方法。该方法首先利用规则抽取技术从论文中抽取具有研究水平信息的句子;然后,构建正则表达式从具有研究水平信息的句子中精确抽取出能够代表单篇学术论文研究水平的数值信息;最后,通过对比分析抽取出的研究水平数值信息进而判断学术论文的创新力。[结果/结论]通过对碳纳米管研究领域的实证研究证明,该方法能够有效、迅速和准确地从论文内容角度对单篇学术论文的创新力进行评价。 [ Purpose/significance] Innovation is the essential requirement of academic paper, and how to effectively evaluate the innovation of academic paper has drawn great attention from domestic and foreign experts and scholars. The development of information technology makes it possible to automatically evaluate a single paper' s innovation based on the paper content. [ Method/ process] This paper proposes a method to evaluate a single paper' s innovation based on the research level. Firstly, the paper uses rule extraction technology to extract sentences which have research level information. Secondly, it constructs regular expression to extract numerical information accurately which represents the research level of paper from sentences with research level information. Lastly, through comparative analysis of the extracted numerical information, the level of paper' s innovation is determined. [ Result/conclusion ] Through an empirical study of carbon nanotube, it is found that this method can evaluate a single paper's innovation from the perspective of the paper's content effectively, quickly and accurately.