：文章应用自然语言处理的方法，基于选定领域的文献的题名、摘要、关键词和全文，识别了研究热点并对其内容效度进行了检验。研究发现：关键词在研究热点识别中内容效度较低，存在一定效度风险；从综合指标来看，基于题名、全文和摘要而识别的研究热点具有相对较高的内容效度；从单项指标（算法）来看，基于不同分析单元而识别研究热点的内容效度具有一定的不稳定性。

Present study conducted a Natural Language Processing procedure to test the content validity of the hot research issues identified by titles, abstracts, keywords and full text of academic papers. The findings include: Firstly, the hot research issues identified by keywords facing a high validity risks. Secondly, the content validity of hot research issues identified by titles are highest compared to other analysis unit from a comprehensive perspective. Thirdly, the hot research issues identified by various units are unstable from perspective of solo-index.