A general framework for defining and extending actual causation using CP-logic
A general framework for defining and extending actual causation using CP-logic
复制标题
DOI：
10.1016/j.ijar.2016.05.008
复制
发表时间：
2016-10-01
影响因子：
3.9
通讯作者：
Vennekens, Joost
中科院分区：
文献类型：
--
作者：
Beckers, Sander;Vennekens, Joost
因果建模领域的一个中心问题是提供实际因果关系的适当定义，即，来定义一个特定事件何时引发了另一个事件。虽然目前的研究包含了许多不同的定义，它是充满了歧义和混乱。我们的研究有两个主要目标。首先，我们希望提供一个明确的方式来比较竞争的定义，并改进它们，使它们可以应用于更多样化的实例，包括非确定性的。为了实现这一点，我们提供了一个一般的，抽象的定义，实际的因果关系，制定的表达语言的CP逻辑（因果概率逻辑）的上下文中。然后，我们将证明，Ned Hall最近的三个定义（最初是为结构模型制定的）和我们自己的定义（直接为CP逻辑制定的）可以被视为这个抽象定义的实例并直接进行比较，这也允许他们处理更广泛的例子。其次，我们的框架允许以另一种方式改进实际因果关系的定义，通过纳入常态的影响。Halpern和希区柯克最近的一篇论文借鉴了关于人们因果判断的实证研究，提出了一个分级的和上下文敏感的实际因果关系概念。我们将他们的方法重新表述为我们抽象定义的概率设置，使我们能够改进它。（C）2016 Elsevier Inc. All rights reserved.
A central problem in the field of causal modelling is to provide a suitable definition of actual causation, i.e., to define when one specific event caused another. Although current research contains many different definitions, it is pervaded with ambiguities and confusion. Our research has two main goals. First, we wish to provide a clear way to compare competing definitions, and improve upon them so that they can be applied to a more diverse range of instances, including non-deterministic ones. To achieve this we provide a general, abstract definition of actual causation, formulated in the context of the expressive language of CP-logic (Causal Probabilistic logic). We will then show that three recent definitions by Ned Hall (originally formulated for structural models) and a definition of our own (formulated for CP-logic directly) can be viewed and directly compared as instantiations of this abstract definition, which also allows them to deal with a broader range of examples. Second, our framework allows for improving on definitions of actual causation in another way, by incorporating the influence of normality. A recent paper by Halpern and Hitchcock draws on empirical research regarding people's causal judgements, to suggest a graded and context-sensitive notion of actual causation. We rephrase their approach into the probabilistic setting of our abstract definition, allowing us to improve it. (C) 2016 Elsevier Inc. All rights reserved.