摘要：以1994 - 2006年Web of Science (含SCIE,SSCI和A&HCI)收录的中国发表论文为样本，从中国参与发表Web of Science文章各学科的总量、速度、及学科分布研究了中国在国际科学研究中的学科地位与学科优势。同时，运用波士顿矩阵方法，按照文章数量与发展速度两个维度将中国参与发表Web of Science文章所涉及的242个学科分为四个类别：“明星”，“奶牛”，“问号”，“瘦狗”，针对不同的学科类别进行了相应的分析并提出各自的发展战略。

Abstract: Based on the papers published with a Chinese address and collected by the Web of Science (including SCI, SSCl, and A&HCI) during the period of 1994-2006, the scientometrics and data mining approach are adoped to analyze disciplinary patterns of Chinese publications, including quantity and growth rate for the different disciplines. In addition, the BCG matrix analysis is used to divide the 242 disciplines into four different categories, that is, stars, cash cows, question marks, and skinny dogs. Policy recommendations are made based on the analysis.