KNOWLEDGE BASED SYSTEM Definition Importance and Types Definition

KNOWLEDGE BASED SYSTEM Definition, Importance and Types

Definition ■ Body of facts and principles accumulated by human kind or act, fact. ■ Having a familiarity with language, concepts, procedures, rules, ideas, abstraction ■ Intelligence requires the possession and access to knowledge.

Types ■ Declarative: – Passive knowledge expressed as facts of the world. – Personnel data in a database is typical of declarative knowledge. ■ Procedural: – Procedural knowledge is complied knowledge related to performance of some task. – Step to solve a complex problem ■ Heuristics: – Knowledge used to make good judgments, tricks.

Other Concepts ■ Beliefs: – Any meaningful and coherent expression that can be represented. – Can be true or false ■ Hypothesis: – A justified belief which is could be true or false. ■ Knowledge as true justified belief. ■ Meta-Knowledge is knowledge about knowledge.

Components of KBS I/o Unit Inference Control Unit Knowledge Base ■ The I/O unit is for input and output of a system ■ Inference control system control the information flow from one unit to another ■ Knowledge base contains all the data from a specific area of field.

Expert System ■ An expert system is a knowledge-based program that provides "expert quality" solutions to problems in a specific domain. ■ knowledge is extracted from human experts ■ knowledge must be encoded into a formal language ■ attempts to imitate expert's methodology and performance ■ both theoretical and practical (heuristic) knowledge included

Expert System Task ■ Interpretation: – forming high-level conclusions from raw data e. g. geological data for oil exploration, mass spectroscopy ■ diagnosis : – determining the cause of malfunctions in complex situations based on observable symptoms e. g. medical diagnosis, car faults ■ design : – designing system configurations to meet certain performance goals while satisfying constraints e. g. XCON designs VAX computer configuration ■ Prediction: – probable consequences of given situations

Expert System Task Cont. …. ■ Planning: – devising a sequence of actions that will achieve a certain goal given certain starting condition e. g. route planning, treatment planning ■ Monitoring: – continuous monitoring of signals and invocation of actions/alarms as appropriate e. g. intensive care of patients ■ Instruction: – detecting and correcting deficiencies in students' understanding of a subject domain ■ Control: – governing the behavior of a complex environment e. g. chemical process control

Expert Systems Needed ■ Benefit justifies development cost – Commercial cost of mistakes (e. g. XCON) – Human life – Mineral exploration ■ Human expertise not available on site – Test drilling sites – Experts leave the company ■ Symbolic reasoning can solve the problem – ES are not good at physical dexterity or perceptual skills (e. g. physical exam. of patient) ■ Problem domain is well-structured and does not require common-sense reasoning ■ Traditional computing methods cannot solve the problem – Generally algorithmic approaches are better than heuristic ones.