Knowledge Representation笔记

Knowledge Representation

认知架构 Cognitive Architectures

Knowledge Level 三要素

• Knowledge Level Hypothesis
  知识为媒介，理性为行为规律（rule）
  There exists a distinct computer systems level, lying immediately above the symbol level, which is characterised by knowledge as the medium and the principle of rationality as the law of behaviour.
• Principle of Rationality
  目标导向性
  If an agent has knowledge that one of its actions will lead to one of its goals, then the agent will select that action.
• Knowledge
  行为根据理性原则来计算
  Whatever can be ascribed to an agent, such that its behaviour can be computed according to the principle of rationality.

Knowledge Representation 知识表达

需要将知识转化为电脑可追溯的方式，比如Reasoning（推理）。推理需要知识和推理。
知识库建立方式之一：Ontologies and Ontology Engineering

Ontology 本体论

(这里我理解的不太好，具体看链接)
An ontology organises everything into a hierarchy of categories.
Ontology意义在于分类，而分类的主要意义在于特点的遗传性。比如一个苹果是水果类的，那么苹果将遗传水果类的可食用性。
Two choices for representing categories in first-order logic: predicates and objects. (不理解)
ontology可用于从知觉输入推断存在对象。对象的话也有很多关系，比如实例，等价，子类，正交关系等。

##################分割线############################

Semantic networks

1. 区分

• object（instance）
• category（class）

2. 区分

• static facts
• transient: 在赋予实例的时候会变化
• default values :can be used as facts in the absence of transient facts
  EXAMPLE:
  • My car is a car (static relationship)
  • A car has four wheels (static attribute)
  • A car’s speed is 0 mph (default attribute)
  • My car is in my garage (default relationship)
  • My car is in High Street (transient relationship)

Description logics

• formal language for constructing and combining category definitions 构造和组合类别定义的形式语言
• efficient algorithms for deciding subset and superset relationships between categories 确定类别之间子集和超集关系

Rule based system

Production rule（产生式规则）: if [condition] then [conclusion]，规则之间具有dependency
同时规则可能面临着不确定性，不确定性的产生主要来源于 不确定的证据，证据和结论之间的不确定联系，模糊的规则

Inference network

由facts（事实）和logic expression（逻辑表达）组成

区分

• Deduction 演绎
  cause + rule => effect
  根据起因找结果，从下向上推理，稳定
  
• Abduction 诱因
  effect + rule => cause
  根据结果找原因，从上向下推理，由于有分支，结果不稳定
  
• Induction 归纳
  cause + effect => rule
  根据现象与结果寻找规律（rules）