关于共被引分析方法的再认识和再思考

Reconsiderations on Co-citation Analysis

共被引（Co-citation）分析方法自1973年提出后便产生了很广泛影响，大家一直以来都按约定俗成的方式去处理共被引矩阵，但是最近对该方法的讨论非常激烈。核心问题围绕“矩阵转化的方法”而展开。在认识他们争论的同时，我们对此也提出了自己的想法和思考。我们从“共被引矩阵作为临近矩阵”这一角度，认为对角线应该采用该作者与其他作者的最大共被引次数＋1来表示；同时我们从稳定性和程序的统一性等方面探讨了欧几里得距离的平方这一标准化算法的优越性。最后我们用实例来证实了自己的想法的可行性和准确性。以期对以后该方法的使用提供参考。

Co-citation has been affecting wildly since put forward in 1973, most of us deal with the co-citation matrix in the ways established by usage, but recently the discussions on it are hot. The core problem is how to convert the matrix. At the same time, we also put forward our thoughts and ideas. We think the diagonal value should be put with maximum ＋ 1 in terms of cocitation matrix as a proximity one. Meanwhile, we discuss the advantages of squared Euclidean distance from the aspects of stability and oneness of the processing program. In the end, we take an example to approve our thoughts and hope these can supply some beneficial reference to further use of this method.