动物识别系统实验 1.基于规则产生式系统结构： 基于规则的产生式系统一般由规则库（知识库）、综合数据库和推理引擎（推理机）三部分组成，规则库中它的基本组成框架如下图1所示。 知识库由谓词演算事实和有关讨论主题的规则构成，综合库又称为上下文，用来暂时存储推理过程中的结论和数据。推理机是用规则进行推理的过程和行为。知识采集系统是领域专家把相关领域的知识表示成一定的形式，并输入到知识库中。解释系统通过用户输入的条件来分析被系统执行的推理结构，并将专家知识以易理解的方式并把知识解释给用户。 2. 简单动物识别产生式系统结构： ⑴ 知识库 ANIMAL的知识库非常小，仅仅包含16条规则（一 般说来，一个产生式系统的知识库应包含≥几百条规则）； ⑵ 解空间很小，仅仅包含8个解，或8个最高假设（在一个特定 的动物园里，共有虎、豹、长颈鹿、斑马、鸵鸟、企鹅和信天 翁、八哥等8种动物）； ⑶ 初始事实集合很小，仅仅包含21个事实，如图中的F1至F21； ⑷ 数据（即事实、证据、断言），知识（即浅层知识，规则）和推 理都是精确的，即确定性的； ⑸ 知识库。-Animal Identification System Experiment 1. Rule-based system architecture: rule-based production system by the rule base (KB), a comprehensive database and inference engine (inference engine) consists of three parts, the rule base of its most basic framework is as followsshown in Figure 1. Knowledge is the rules of the facts and the topics discussed by the predicate calculus, the comprehensive library also known as context to conclusions and data is temporarily stored in the reasoning process. The inference engine is the process of inference rules and behavior. Knowledge acquisition system is the domain expert knowledge of the related fields into definite form, and entered into the knowledge base. Interpretation system conditions input by the user the inference structure analysis was performed by the system, and the expert knowledge is easy to understand and explain the knowledge to the user. 2 simple animal identification system structure: (1) Knowledge Base ANIMAL knowledge base is