基于Rao-Stirling指数和LDA模型的领域学科交叉主题识别——以纳米科技为例

Interdisciplinary Literature Discovery Based on Rao-Stirling Diversity Indices: Case Studies in Nanoscience and Nanotechnology

【目的/意义】基于Rao-Stirling指数和LDA模型进行领域学科交叉主题识别,并以纳米科技为例验证将Rao-Stirling指数和LDA模型用于领域学科交叉主题识别的有效性和适用性。【方法/过程】基于Rao-Stirling指数测度领域文献学科交叉程度,设定阈值发现高度学科交叉文献。基于LDA模型对筛选出的学科交叉文献进行主题识别,发现学科交叉点和学科交叉研究主题。【结果/结论】基于Rao-Stirling指数从引文的角度进行领域文献学科交叉测度可以有效地发现与某领域相关的学科交叉文献,且有利于大数据集的学科交叉文献发现研究的实现。基于LDA模型进行学科交叉主题识别可以有效地发现学科交叉主题。两方法的组合应用为发现某领域学科交叉主题研究提供一种新视角。

【Purpose/significance】This paper aims to identify interdisciplinary topic based on Rao-Stirling diversity indices and LDA(Latent dirichlet allocation)model,the field of nanoscience and nanotechnology was taken as the example to verify the method’s efficiency and applicability【Method/process】Measuring the interdisciplinary extent of domain publications based on Rao-Stirling diversity indices to find domain-related interdisciplinary publications by setting up a threshold.Mining the topic of domain-related interdisciplinary publications based on LDA model to discover domain intersections.【Result/conclusion】The citation analysis of interdisciplinary degree of literature based on the Rao-Stirling Index can effectively find interdisciplinary literature related to a certain field,and the algorithm is of low complexity,which is beneficial to the discovery of interdisciplinary literature of large data sets.The interdisciplinary topic identification based on LDA model can effectively discover interdisciplinary themes of a certain domain.The combination of two methods provides a new perspective for domain interdisciplinary topic identification study.