泛逻辑学理论——机制主义人工智能理论的逻辑基础

PDF下载HTML

Title:

Universal logic theory: logical foundation of mechanism-based artificial intelligence theory

作者:

何华灿

西北工业大学 计算机学院, 陕西 西安 710072

Author(s):

HE Huacan

School of Computer Science, Northwestern Polytechnical University, Xi’an 710072, China

关键词:

泛逻辑学; 机制主义人工智能; 不确定性; 柔性神经元; 智能信息处理算子库

Keywords:

universal logic; mechanism-based artificial intelligence theory; uncertainty; flexible logic; intelligent information processing operator library

分类号:

TP18

DOI:

10.11992/tis.201711033

摘要:

当前，世界各主要大国都把人工智能作为它们的国家战略。人工智能的发展正在快速改变着人类的生活方式和思想观念。在中国，有一小批研究者20多年来一直在基于辩证唯物主义潜心研究具有普适性的人工智能基础理论，包括智能的形成机制、逻辑基础、数学基础、协调机理、矛盾转化等。终于，他们各自建立了机制主义人工智能理论、泛逻辑学理论、因素空间理论、协调学、可拓学、集对分析等。其中，机制主义人工智能理论是基于智能形成机制的通用理论，它能把现有的结构主义、功能主义和行为主义三大流派有机地统一起来，使意识、情感、理智成为三位一体的关系；因素空间理论是机制主义人工智能理论的数学基础；泛逻辑学理论是机制主义人工智能理论的逻辑基础。本文介绍了泛逻辑学理论的基本思想、理论基础和应用方法，阐明它的理论意义和应用价值。特别需要指出的是，在广义概率论基础上建立的命题泛逻辑（包括刚性逻辑和柔性逻辑），可看成一个完整的命题级智能信息处理算子库，库中完整地包含了全部18种柔性信息处理模式（包括16种布尔信息处理模式），可用类型编码<a,b,e>来严格区分，用它可寻找到适合自己的信息处理算子完整簇来使用。在每一个信息处理模式中，各种不确定性的组合状态由不确定性程度属性编码<k,h,β,e>来严格区分，用它可在本信息处理模式的算子完整簇中精确选择具体的算子来使用。这表明柔性信息处理本质上是一把密码锁，它需要专门的密码<a,b,e>+<k,h,β,e>才能正常打开，不能乱点鸳鸯谱。通过只有18种模式，每种模式可以从最大算子连续变化到最小算子，已经证明了没有一个命题算子被遗漏。

Abstract:

Recently, the world’s major countries have begun to consider artificial intelligence (AI) as part of their national strategies. The widespread development of AI is changing our lives and ideas. In China, a small number of researchers have been concentrating their attentions onto the fundamental theory of intelligence based on dialectical materialism for more than 20 years, including the formation mechanism of intelligence, logical basis, mathematical foundation, coordination mechanism and contradiction transformation.At last, they set up the mechanism-based artificial intelligence theory, the universal logic theory, the factor space theory, the coordination theory, the extenics theory and the analysis of set pairs. The mechanism-based artificial intelligence theory is the general theory of intelligence based on the formation mechanism, it unifies the strucrualism, functionalism, and behaviorism, the existing three isolated major schools of artificial intelligence, into a harmonious unity, and makes the primary consciousness, emotion, and lntellect into a trinity; factor space theory is the mathematical foundation for mechanism-based artificial intelligence theory; universal logic theory is the logical based for mechanism-based artificial intelligence theory. In this paper, we introduces the basic theory and application of universal logic theory and clarifies its significance and value. Propositional universal logic (containing both hard and flexible logics) is based on “generalized probability theory”, and can be regarded as a proposition-level library of intelligent information processing operators. It contains all 18 flexible information processing modes (including 16 Boolean information processing modes). The operators are strictly expressed as <a,b,e>, so appropriate complete clusters of information processing operators can be identified. In each mode, different uncertainties are determined by the degree of encoding <k,h,β,e>, which connects with the direct operator required. Consequently, flexible information processing can considered to be a coded lock, with <a,b,e> + <k,h,β,e> as its passwords. Additionally, all of the passwords are saved in the library, thereby satisfying the demand for “one lock-one passwords”.