哲学社会科学版
陕西师范大学学报(哲学社会科学版)
哲学研究
人工智能视野下的知识体系修正理论
PDF下载 ()
任晓明, 林艺霏
(南开大学 哲学院, 天津 300350)
任晓明,男,四川泸州人,四川大学文科讲席教授,南开大学哲学院教授,博士研究生导师。
摘要:
从人工智能的视角看,知识体系修正具体表现为信念修正。信念修正理论的发展史就是一部与人工智能分支融合共同进步和发展的历史。信念修正理论可以为人工智能提供思想方法,同时人工智能方法也能为信念修正提供动力支持。这种跨域研究对人工智能和信念修正理论研究来说是相得益彰的。人工智能不断试错的方法是推动信念修正理论发展的“推进器”。人工智能方法在一定程度上是“实效主义”的。它没有预设固定的理论模式,只要能解决问题,无论何种方法、路线都可以尝试,这种不断“试错”的动态可修正方法促进了信念修正理论的发展,使该理论不断突破局限,为解决科学和哲学问题开辟了新的思路。
关键词:
知识体系; 信念修正; 人工智能; AGM理论; 计算树逻辑模型
收稿日期:
2021-10-20
中图分类号:
T918; G642
文献标识码:
A
文章编号:
1672-4283(2022)02-0121-08
基金项目:
国家社会科学基金重大项目“现代归纳逻辑的新发展、理论前沿与应用研究”(15ZDB018)
Doi:
10.15983/j.cnki.sxss.2022.0311
Knowledge System Revision From the Perspective of Artificial Intelligence
REN Xiaoming, LIN Yifei
(College of Philosophy, Nankai University, Tianjin 300350)
Abstract:
From the perspective of artificial intelligence,knowledge system revision is embodied in belife revision.The development history of belief revision is a history of common progress and development with the branches of artificial intelligence. Belief revision can provide artificial intelligence with ideas and methods, and at the same time, artificial intelligence methods can also provide motivational support for belief revision. This kind of crossdomain research is complementary to artificial intelligence research and belief revision theory research. The continuous trial and error method of artificial intelligence is the “propeller” to promote the development of belief revision.The artificial intelligence method is “pragmatic” to a certain extent, because it does not presuppose a fixed theoretical model, as long as the problem can be solved, no matter what method or route can be tried. This dynamic and modifiable method of continuous “trial and error” promotes the development of belief revision, which breaks through its limitations and opens up new ideas for solving problems.
KeyWords:
knowledge system; belief revision; artificial intelligence; AGM theory; computational logic