基于主题和指标特征的研究前沿识别系统设计与实现

Design and Implementation of a Research Frontier Identification System Based on Theme and Index Characteristics

准确、快速地识别研究前沿,对于促进科技创新、攻克关键技术、推动学科发展和解决重大问题具有重要意义。以基金项目、学术论文资源为基础,利用LDA主题模型、BERT模型、Word2Vec等方法对科技资源进行主题内容挖掘,同时从新兴度、创新性、交叉性、关注度和中心性5个维度,构建能够表征和识别前沿主题的指标体系,并研发研究前沿识别与多维分析系统。该研究可为更加科学、准确、前瞻地识别科学研究前沿、分析演化路径提供具有应用价值的方法与工具。

Accurately and quickly identifying research frontiers is of great significance for promoting technological innovation,grasping key technologies,promoting disciplinary progress,and solving major problems.Based on fund projects and academic paper resources,this paper uses the LDA topic model,BERT model,Word2Vec,and other methods to mine the subject content of scientific and technological resources.At the same time,from the five dimensions of emerging,innovative,interdisciplinary,attention,and centrality,this paper constructs an index system that can characterize and identify cutting-edge themes,and develops a cutting-edge identification and multidimensional analysis system as well.This study provides valuable methods and tools for identifying scientific research frontiers,analyzing evolutionary paths,and providing more scientific,accurate,and forward-looking insights.