A Preliminary Logic-based Approach for Explanation Generation

A Preliminary Logic-based Approach for Explanation Generation
复制标题

一种基于逻辑的初步解释生成方法

DOI:
--
复制
发表时间:
2019
期刊:
影响因子:
--
通讯作者:
S. Vasileiou
S. Vasileiou
中科院分区:
--
文献类型:
--
作者:
S. Vasileiou

文献摘要

被引文献

相似文献

在解释生成问题中,一个智能体需要识别并向另一个智能体解释其决策的原因。该领域的现有工作主要局限于基于规划的系统,这些系统使用自动化规划方法来解决问题。在本文中,我们从一个新的角度来处理这个问题,提出了一个用于解释生成的通用的基于逻辑的框架。特别是,给定一个包含公式 φ 的知识库 KB1 和一个不包含 φ 的第二知识库 KB2,我们寻求识别一个解释,该解释是 KB1 的子集,使得 KB2 和包含 φ 的并集。我们定义两种类型的解释,模型解释和证明理论解释,并使用成本函数来反映解释之间的偏好。此外,我们提出了为命题逻辑实现的算法,该算法计算此类解释并在随机知识库中凭经验对其进行评估,并且
In an explanation generation problem, an agent needs to identify and explain the reasons for its decisions to another agent. Existing work in this area is mostly confined to planning-based systems that use automated planning approaches to solve the problem. In this paper, we approach this problem from a new perspective, where we propose a general logic-based framework for explanation generation. In particular, given a knowledge base KB1 that entails a formula φ and a second knowledge base KB2 that does not entail φ, we seek to identify an explanation that is a subset ofKB1 such that the union ofKB2 and entails φ. We define two types of explanations, modeland proof-theoretic explanations, and use cost functions to reflect preferences between explanations. Further, we present our algorithm implemented for propositional logic that compute such explanations and empirically evaluate it in random knowledge bases and