The aim of this study is to introduce two groups of impact indicators, Weighted Altmetric Impact (WAI) and Inverse Altmetric Impact (IAI). WAI is based in weights from the contributions of each metric to different components or impact dimensions (Principal Component Analysis). IAI is calculated according to the inverse prevalence of each metric in different impact dimensions (TF/IDF). These indicators were tested against 29,500 articles, using metrics from Altmetric.com, PlumX and CED. Altmetric Attention Score (AAScore) was also obtained to compare the resulting scores. Several statistical analyses were applied to value the advantages and limitations of these indicators. Frequency distributions showed that each group of metrics (Scientific Impact, Media Impact and Usage Impact) follows power law trends although with particular patterns. Correlation matrices have depicted associations between metrics and indicators. Multidimensional scaling (MDS) has plotted these interactions to visualize distances between indicators and metrics in each dimension. The 2018 Altmetric Top 100 was used to distinguish differences between rankings from AAScore and the proposed indicators. The paper concludes that the theoretical assumptions of dimensions and prevalence are suitable criteria to design transparent and reproducible impact indicators.

这项研究的目的是介绍两组影响指标：加权高度影响（WAI）和逆高度影响（IAI）。WAI的权重基于每个指标对不同组件或影响维度的贡献（主组件分析）。根据每个指标在不同影响维度（TF / IDF）中的逆向流行度来计算IAI。这些指标已使用Altmetric.com，PlumX和CED的指标针对29,500篇文章进行了测试。还获得了高度注意力集中分数（AAScore）来比较结果分数。进行了一些统计分析，以评估这些指标的优缺点。频率分布表明，每组度量标准（科学影响，媒体影响和使用影响）都遵循幂律趋势，尽管具有特定的模式。相关矩阵描述了指标和指标之间的关联。多维标度（MDS）绘制了这些交互作用，以可视化每个维度中指标和指标之间的距离。2018年Altmetric前100名用于区分AAScore排名与拟议指标之间的差异。本文得出的结论是，关于尺寸和流行程度的理论假设是设计透明且可重现的影响指标的合适标准。