SHAPE-ID Literature Review dataset: subject co-occurrence matrix

Background and methodology: The subject co-occurrence matrix represents the pairs of All Science Journal Classification (ASJC) disciplines that co-occur in journals represented in the SHAPE-ID Literature Review dataset, prepared for the purposes of quantitative analysis. We take disciplinary affiliations of journals as a proxy of disciplinary characteristics of the journal articles in the Literature Review dataset, mindful of the fact that a particular article might deviate from the disciplinary affiliation of the journal in which it was published. However, since there was no data readily available on item level, and manual disciplinary encoding of all the items in the bibliography was beyond the scope of this study, the method used is the best approximation of the presence of discourse on interdisciplinarity and transdisciplinarity, in and between disciplines. In the matrix, each co-occurrence value is weighted by the number of journals that feature the given pair of disciplines, and by the number of articles represented in the dataset that feature in these journals. E.g. if Journals J1 and J2 each featured disciplines D1 and D2, and if 4 articles from J1 and 7 articles from J2 are represented in the SHAPE-ID Literature Review dataset, the co-occurrence value is 11. The pairings cross-referencing a single discipline (e.g. 1202 History in both first row and first column) correspond to the co-occurence value of mono-disciplinary journals. Description of the file: This is a csv file containing a 308x308 cell matrix with ASJC disciplines in first rows and columns, and co-occurrence value in the remaining cells.

背景与研究方法：主题共现矩阵（subject co-occurrence matrix）呈现的是SHAPE-ID综述数据集（SHAPE-ID Literature Review dataset）收录期刊中，两两同时出现的《所有学科期刊分类法》（All Science Journal Classification, ASJC）学科组合，本数据集专为定量分析而构建。我们以期刊的学科归属作为该综述数据集内期刊文章学科特征的代理指标，同时注意到单篇文章可能与其刊载期刊的学科归属存在偏差。但由于无法直接获取条目级数据，且对参考文献库中所有条目进行人工学科编码超出了本研究的范畴，因此本方法已是对学科内部及学科间跨学科性与超学科性话语存在性的最佳近似。在该矩阵中，每一组共现的权重由同时包含该两组学科的期刊数量，以及这些期刊内收录于本数据集的文章总数共同决定。例如：若期刊J1与J2均涵盖学科D1与D2，且SHAPE-ID综述数据集内收录有J1的4篇文章与J2的7篇文章，则该学科组合的共现值为11。仅涉及单一学科的交叉组合（例如首行与首列均为1202 历史学）对应的共现值，代表单学科期刊的相关数值。文件说明：本文件为CSV格式，包含一个308×308的单元格矩阵，矩阵的首行与首列为ASJC学科名称，其余单元格则填充对应学科组合的共现值。