Knowledge-Based Reasoning

Knowledge-Based Reasoning (KBR) is a fundamental aspect of Artificial Intelligence that

enables systems to use stored knowledge to infer new information, make decisions, and solve
problems. Unlike purely data-driven approaches, KBR relies on structured knowledge
representations and logical inference to derive conclusions.

Knowledge-Based Agents:
A Knowledge-Based Agent (KBA) is an intelligent system that utilizes a structured Knowledge
Base and an Inference Engine to make decisions, reason about the world, and take appropriate
actions.

Components of a Knowledge-Based Agent:

(a) Knowledge Base (KB): A repository of structured information containing facts, rules, and
domain knowledge. Can be stored using First-Order Logic (FOL), Semantic Networks,
Ontologies, etc.

Example: For "All humans are mortal" and "Socrates is a human", the logical inference will be
"Socrates is mortal".

(b) Inference Engine: The reasoning mechanism that derives new facts from the knowledge
base. Uses deductive reasoning, inductive reasoning, probabilistic reasoning, and rule-based
inference.

Example: IF (Symptom = Fever) AND (Symptom = Cough) THEN (Possible Diagnosis = Flu).

(c) Perception Module: Collects data from sensors, databases, or human inputs.

Example: In a self-driving car, the perception module collects data from cameras, LiDAR, and
GPS.

(d) Action Module: Takes appropriate actions based on the agent's reasoning.

Example: A robotic arm detects a defective product and removes it from an assembly line.

(e) Learning Mechanism: Updates the knowledge base using machine learning, feedback
mechanisms, and experience.

Example: A medical diagnosis system updates its knowledge as new diseases are discovered.
Knowledge-Based Agent Model:

Mathematically, a knowledge-based agent can be expressed as:

KBt = KBt−1 + Perceptt

Actiont = Reasoning(KBt)

Where KBt is the updated knowledge base at time t, Perceptt is the new information received,
and Reasoning(KBt) applies inference mechanisms to decide on the next action.

Types of Knowledge-Based Agents:

(a) Logic-Based Agents: It uses formal logic to represent knowledge and make decisions. It
utilizes propositional logic or first-order logic to reason systematically.

Example: Expert systems in healthcare, such as MYCIN, which diagnoses bacterial infections
using logical rules.

(b) Rule-Based Agents: It is an AI system that makes decisions using a set of predefined rules.
These rules follow the IF-THEN logic, allowing the agent to act based on observed conditions.

Example: Chatbots and recommendation systems.

(c) Learning-Based Agents: It improves its performance over time by learning from
experience, data, or interactions with the environment. It uses machine learning to improve
their knowledge base over time.

Example: AlphaGo, which learns Go strategies by playing millions of games.

(d) Hybrid Agents: It combines multiple reasoning and learning techniques to achieve better
decision-making and adaptability. It integrates rule-based reasoning, logic-based reasoning,
and machine learning to leverage each approach's strengths while overcoming their limitations.

Example: IBM Watson, which integrates knowledge bases with deep learning for medical
diagnosis.

Real-World Examples of Knowledge-Based Agents:

(a) Expert Systems in Medicine: Uses medical databases, research papers, and patient
symptoms to recommend treatments.

Example: IBM Watson for Oncology

(b) Autonomous Vehicles: Knowledge base includes traffic laws, sensor inputs, and maps to
make real-time driving decisions.

Example: Self-Driving Cars (Tesla, Waymo)

(c) Virtual Assistants: Uses knowledge bases and natural language processing to answer
queries.

Example: Siri, Alexa, Google Assistant

(d) Fraud Detection Systems: Uses rule-based reasoning and machine learning to detect
fraudulent transactions.

Example: Credit Card Fraud Detection (Mastercard, Visa AI Systems)

Facets of Knowledge:
AI systems rely on different types of knowledge to function efficiently. Each type serves a
specific role in reasoning, decision-making, and problem-solving.

(a) Declarative Knowledge (Descriptive Knowledge): It consists of facts and information

about the world that AI systems store and retrieve when needed. It represents "what" is known
rather than "how" to do something.

Example: ‘Delhi is the capital of India’, ‘Bhubaneswar is a city in Odisha’, etc.

(b) Procedural Knowledge (How-To Knowledge): It defines the steps or methods required
to perform specific tasks. It represents "how" to accomplish something rather than just stating
a fact.

Example: ‘How to solve a quadratic equation?’, ‘How to drive a car?’, etc.

(c) Meta-Knowledge (Knowledge About Knowledge): It refers to knowledge about how

information is structured, used, and validated. It helps AI determine the reliability, relevance,
and applicability of knowledge in different scenarios. It is used for filtering misinformation,
optimizing learning strategies, and improving decision-making.

Example: Deciding whether a piece of medical advice comes from a trusted scientific source
or a random blog post.

(d) Heuristic Knowledge (Experience-Based Knowledge): It is derived from experience,

intuition, and trial-and-error methods. It allows AI systems to make educated guesses or
approximate solutions when exact answers are difficult to compute.

Example: A navigation system suggesting an alternate route based on past traffic patterns.
(e) Common-Sense Knowledge: It represents basic understanding about the world that
humans acquire naturally but is challenging for AI to learn.

Example: ‘Gravity pulls objects downward.’, ‘Water is wet, and fire is hot.’, ‘If you drop
something, it will fall.’

(f) Domain-Specific Knowledge: It focuses on specialized fields such as medicine, finance,

law, or engineering. It includes highly detailed and structured information relevant to a
particular industry.

Example: In the medical field, AI-driven diagnostic systems rely on knowledge about
symptoms, diseases, and treatments. Similarly, financial AI models use economic indicators,
risk assessments, and market trends.

Apart from these, several other types of knowledge are utilized by AI systems. By integrating
different facets of knowledge and inference mechanisms, AI systems can become more
intelligent, interpretable, and capable of handling complex decision-making tasks.