knowledge representation systems

• Artificial intelligent systems usually consist of various components to display their intelligent behavior
• Components of an AI system

  • Perception
  • Learning
  • Knowledge Representation & Reasoning
  • Planning
  • Execution

  
• Perception Component


  Retrieves data or information from the environment
• Learning Component


  Learns from the captured data by the perception component
• Knowledge Representation and Reasoning


  Shows the human-like intelligence in the machines
• Planning Component


  Gives an initial state, finds their preconditions and effects, and a sequence of actions to achieve a state in which a particular goal holds
• Execution Component


  Executes the entire process planned by the Planning component
• Knowledge plays a vital role in intelligence as well as creating artificial intelligence
• Knowledge demonstrates the intelligent behaviour in AI agents or systems
• An agent or system can act accurately on some input only when it has the knowledge or experience about the input
• Techniques of Knowledge Representation in AI

  • Logical Representation
  • Semantic Network Representation
  • Frame Representation
  • Production Rules

  
• Logical Representation


  • Helps to perform logical reasoning
  • Basis for programming languages
• Logical Representation


  • Has some restrictions and is challenging to work with
  • May not be very natural, and inference may not be very efficient
• Semantic Network Representation


  • Natural representation of knowledge
  • Conveys meaning transparently
  • Simple and easy to understand
• Semantic Network Representation


  • Takes more computational time at runtime
  • Inadequate as they do not have any equivalent quantifiers
  • Not intelligent and depend on the creator of the system
• Frame Representation


  • Makes programming easier by grouping related data
  • Easy to understand and visualize
  • Easy to add slots for new attributes and relations
  • Easy to include default data and search for missing values
• Frame Representation


  • Inference mechanism cannot be smoothly proceeded
  • Very generalized approach
• Production Rules


  • Expressed in natural language
  • Highly modular and can be easily removed or modified
• Production Rules


  • Does not exhibit any learning capabilities and does not store the result of the problem for future uses
  • Many rules may be active, making it inefficient
• Representation Requirements


  • Representational Accuracy
  • Inferential Adequacy
  • Inferential Efficiency
  • Acquisitional efficiency
• Approaches to Knowledge Representation in AI

  • Simple Relational Knowledge
  • Inheritable Knowledge
  • Inferential Knowledge
• Logic has been called the calculus of computer science, because logic plays a fundamental role in computer science, starting from the construction of computers to the computing devices beyond our ability to construct, such as computer architecture (digital gates, hardware verification), software engineering (specification, verification), programming languages (semantics, type theory, abstract data types, object-oriented programming), databases (relational algebra), artificial intelligence (automated theorem proving, knowledge representation), algorithms and theory of computation (complexity, computability), etc.
• Logic also plays important roles in other fields, such as mathematics and philosophy. In mathematics, logic includes both the mathematical study of logic and the applications of formal logic to other areas of mathematics.
• In artificial intelligence, we need intelligent computers which can create new logic from old logic or by evidence, so generating conclusions from evidence and facts is termed as Inference.
• Inference rules are the templates for generating valid arguments.
• Inference rules are applied to derive proofs in artificial intelligence, and the proof is a sequence of the conclusion that leads to the desired goal.
• Proposition


  A statement or assertion that expresses a judgment or opinion
• Every statement can be either true or false. True or false are called the truth values of a statement.
• Fields of logic


  • Set theory
  • Model theory
  • Proof theory
  • Computability theory
  • Fuzzy Set Theory
• Formal logic


  Formal logic has tended to be termed either philosophy of logic or philosophical logic
• Philosophical logic


  Deals with arguments in the natural language used by humans
• Logical operators/connectives


  Words such as "not" (negation), "and" (conjunction), "or" (disjunction), "implies" (implication), "if-then"
• Propositional variables


  Symbols like p, q, r, s, t that denote simple or composed statements
• Propositional logic provides a good start at describing the general principles of logical reasoning, but it is not possible to express the general properties of many important sets, especially when a set is infinite.
• First-order logic is a considerably richer logic than propositional logic.
• First-order logic


  Logic in which the predicate of a sentence or statement can only refer to a single subject
• Subject


  The main part of a first-order logic statement
• Predicate


  A relation that binds two atoms together in a first-order logic statement
• Rule-based expert system


  The simplest form of artificial intelligence that uses prescribed knowledge-based rules to solve a problem
• Rule-based systems (production systems or expert systems) are the simplest form of artificial intelligence.