学术论文的数量特征与文本趋势

结合文献计量与文本挖掘方法，以《统计研究》自1984-2013三十年载文为主要研究样本，探讨学术论文的数量特征与文本趋势。研究结果表明，（1）统计研究领域与知识结构的变迁大致经历了经济统计、实证方法与经济统计、模型化和经验数据实证方法等几个主题；（2）无论是作者、机构还是基金资助，发文量皆基本符合“二八原则”；（3）并非近十年左右，而是近三年的论文更频繁地被引；（4）声誉是一个长期概念，因此，作为测度学术声誉和影响力的重要指标，影响因子的趋势相比于某一年份的实际值更为紧要。

Quantitative Characters and Structural Change of Academic Papers

Based on bibliometrics and text mining, with the articles published in "Statistical Research (1984-2013)" as samples, we study the quantitative characters and structural change of academic papers. The results show that (1) knowledge structure has experienced some stages like economic statistics, empirical method and economic statistics, modeling and etc.; (2) the number of papers (authors, institutions, or fund) are largely in line with the Matthew effect; (3) the past nearly three years of papers are more frequently cited; (4) as an important indicator to measure the academic reputation, the positive and stable trend of impact factor is more important, because reputation is a long-term concept.