A Computational Theory of Discovery

归纳推理中使用的模型植根于讨论科学方法的科学哲学家所使用的模型。那里的目标，以及学习理论中先前的工作，都是为了对正在考虑的现象做出解释。然而，科学家很少直接致力于完整解释的宏伟目标。人们追求的是寻找有关观察数据的特征和事实的更小目标。将研究各种类型的算法，可以被解释为发现其最终结果。我们建议考虑发现其预期结果的计算。将开发和研究发现逻辑。这项研究与当代科学特别相关，因为自动数据生成技术产生了足够的数据量，以压倒人类的分析能力。我们的工作的目标是准确地阐明自动数据分析算法可以完成什么和不能完成什么。这些算法被用于万维网搜索引擎的数据挖掘和文本分析。

he models used in inductive inference have theirroots in the models used by the philosophers ofscience who were discussing the scientific method.The goal there, and in prior work in learning theory,was to come up with an explanation of the phenomenonunder consideration.However, scientists rarely work directly for the grand goalof a complete explanation.The more modest goal of finding features andfacts about the observed data is pursued.A variety of types of algorithms that could be construed asdiscovering their final result will be investigated.We propose to consider computations that discover rather than computetheir intended result.A logic of discovery will be developed and investigated.This study is particularly relevant to contemporaryscience as automated data generation techniquesproduce sufficient volumes of data tooverwhelm the analysis abilities of humans.The goal of our work is to illuminateprecisely what can and cannot be accomplishedby automatic data analysis algorithms.Such algorithms are used in data mining andtext analysis for world wide web search engines.