OpenMAPP: open meta-analysis in particle physics

现代粒子物理实验是通过巨大的国际设施和合作进行的，产生了数百篇针对无数类型测量的科学论文。特别是大型强子对撞机十多年的活动，并没有为新物理的任何“简单”假设模型提供强有力的统计证据，但这并不意味着这些研究一文不值--相反，它们的全部能量并没有被充分利用，因为它们被孤立地考虑。如果我们能将它们结合在一起，我们应该能够对新的物理模型进行更强有力的研究，包括比最初激励每项研究的模型更复杂的研究。不幸的是，现有的研究并没有以一种单一的格式收集在一起，以重新利用：它们以不同的方式存储，跨越不同的系统和数据库，并带有为在不同工具包中使用而设计的基本数据和代码。OpenMAPP项目将在这些系统之间建立必要的联系，通过扩展Web应用程序、控制模拟并以通用统计格式收集预测，以及找出如何以一致的方式组合测量结果，来实现它们的结合。这些系统还将提供一个生态系统，在其中可以比以前更好地设计和存储新的实验研究，因此未来的粒子物理总是可以与过去的测量相结合使用。

Modern particle physics experiments are conducted via huge international facilities and collaborations, resulting in hundreds of scientific papers aimed at myriad types of measurement. More than a decade of activity at the Large Hadron Collider in particular has failed to show up strong statistical evidence for any "simple" hypothetical models of new physics, but that does not mean that these studies have been worthless -- rather, their full power has not been exploited because they have been considered in isolation. If we could combine them all together, we should be able to make much stronger investigations of new-physics models, including more complex ones far beyond those which originally motivated each study.Unfortunately, the existing studies are not gathered together in a single format for repurposing: they are stored in different ways, across different systems and databases, and with their essential data and code designed for use in different toolkits. The OpenMAPP project will build the necessary links between these systems to allow their combination, by extending Web applications, controlling simulations and gathering predictions in a common statistical format, and figuring out how to combine measurements in a consistent way. These systems will also provide an ecosystem within which new experimental studies can be designed and stored better than before, so the particle physics of the future can always be used in combination with measurements from the past.