规划文本和论文数据对比视域下细主题粒度科学研究前沿识别研究——以碳纳米管研究领域为例

Scientific Research Front Identification of Fine-grained Topic from the Comparison of Planning Texts and Papers: A Case Study of Carbon Nanotubes

[目的/意义]文章利用规划文本和论文数据两种数据源,提出一种基于多维度划分的细主题粒度研究前沿识别方法。[方法/过程]获取碳纳米管研究领域相关规划文本和论文,将两种数据源分别划分为理论创新、实际应用和风险管理三个维度,在各个维度内利用主题模型和回溯原文方法进行细粒度主题抽取,通过不同数据源不同维度细粒度主题对比分析,识别相应研究前沿主题。[结果/结论]实验结果表明该方法可以更有效地识别出细分领域科学研究前沿主题。

[ Purpose/significance ] Using the data sources of planning texts and paper data, this paper proposes a method to i- dentify the fine-grained topics of research fronts based on multidimensional partition. [ Method/process ] First, the paper obtains the planning texts and papers which are related to the carbon nanotubes and divides the two data sources into three dimensions re- spectively, namely theoretical innovation, practical application, and risk management. The topic model and original text retrieval method are used in each dimension to extract the fine-grained topics. Through the comparison and analysis of different fine-grained topics in different dimensions from different data sources, the corresponding front topics are identified. [ Result/conclusion ] The empirical study shows that this method can effectively identify the fine-grained topics of research fronts in the subdivision fields.