Systematic literature reviews (SLR) are commonly undertaken by researchers to stay informed of the latest development in a particular topic, but this manual process is demanding and can only locate and analyze a limited number of articles. We propose a data analytic-based SLR protocol and a set of semi-automated tools to leverage the latest advances in data analytics and facilitate a more effective, objective, and comprehensive SLR process. Our protocol incorporates scraping tools to collect articles from seven bibliographic databases, and text analytics, social network analysis, natural language processing, citation analysis, and main path analysis to analyze a large number of articles. To demonstrate its utility of, we apply the protocol on the topic of “information diffusion in social networks”. The results reveal 11 latent topics under this broad domain along with the most critical articles for each topic, and the connections among the associated 1,229 articles and their references.

研究人员通常会进行系统文献综述（SLR），以了解特定主题的最新进展，但这种手动过程要求很高，而且只能定位和分析有限数量的文章。我们提出了基于数据分析的 SLR 协议和一套半自动化工具，以利用数据分析的最新进展并促进更有效、客观和全面的 SLR 流程。我们的协议包含抓取工具来从七个书目数据库收集文章，以及文本分析、社交网络分析、自然语言处理、引文分析和主路径分析来分析大量文章。为了证明其实用性，我们将协议应用于“社交网络中的信息传播”主题。