27
 FRAMES By: Abisekh Bahadur Niraula (071/MSCSKE/651) Rasu Shrestha (071/MSCSKE/662) Shikhar Basnet (071/MSCSKE/666) Umesh Timalsina (071/MSCSKE/669) Department of Electronics and Computer Engineering IOE, Pulchowk Campus

Frames Artificial Intellegence

Embed Size (px)

DESCRIPTION

Frames are knowledge representation techniques frequently used in AI, knowledge engineering. The concept is similar to inheritance in an OOP paradigm.

Citation preview

  • FRAMES

    By:

    Abisekh Bahadur Niraula (071/MSCSKE/651)

    Rasu Shrestha (071/MSCSKE/662)

    Shikhar Basnet (071/MSCSKE/666)

    Umesh Timalsina (071/MSCSKE/669)

    Department of Electronics and Computer Engineering

    IOE, Pulchowk Campus

  • 1. FRAMES- INTRODUCTION:o Knowledge Representation Technique

    o Similar to Semantic Nets.

    o Proposed by: Marvin Minsky in 1974.

    o Article: A Framework For Representing Knowledge.

    o Common Sense System: Not Rule Based.

    o AIs Primary Data Structures.

    2

  • 1. FRAMES- INTRODUCTION-II:o Frame: Prototype of a Concept

    o Denoting attributes of the concept

    o The class of objects or concepts to which the concept inquestion belongs.

    o And some other things

    o An instance of a Frame.

    o Representation of a specific object.

    o Semantic Nets with Attributes.

    o Description: Similar to Structure Data Type in C.

    3

  • 1. FRAMES- INTRODUCTION-III:o A frame represents an entity as a set of slots

    (attributes) and associated values

    o Represents:

    o General Concept (Frame)

    o Specific Entry (Instance of the Frame)

    o Implicit Relation with other Frames.

    o Concept: Similar to OOS(Object Oriented Systems)

    o Hierarchical Representations: Super and Sub Classes.

    o Classes and Frames: Used interchangebly.

    4

  • 2. EXAMPLES: 5

  • 2. EXAMPLES-HIERARCHY: 6

  • 2. EXAMPLES-HIERARCHY: 7

  • 2. EXAMPLES- I:o FIDO is a Dog.

    o Dog is a concept.

    o Fido is a Particular dog.

    How Can We Represent the Concept/ Notion of theCreature Dog?

    8

  • 2. EXAMPLES:o FIDO is a Dog.

    o Dog is a concept.

    o Fido is a Particular dog.

    How Can We Represent the Concept/ Notion of theCreature Dog?

    A Way: ANIMALS>>>MAMMALS>>DOG

    o FIDO: an instance of the Frame Dog.

    o Notion of Frames: Felt by people out of CommonSense.

    9

  • 2. EXAMPLES- FRAME: 10

    SLOT NAME SLOT VALUES(Default)

    IS_A Mammal

    No_of_legs 4

    Has_Tail Yes

    Color_of_Skin

    Owner

    Type_of_Teeth Sharp

    DOG:

  • 2. EXAMPLES- FRAME: 11

    SLOT NAME SLOT VALUES

    IS A Mammal

    No_of_legs 4

    Has Tail Yes

    Color of Skin White

    Owner Shikhar

    Type of Teeth Sharp

    FIDO:

  • 2. EXAMPLES- FRAME: 12A Book Frame:

    Book Frame

    Slot Filler

    Title AI. A modern Approach

    Author Russell & Norvig

    Year 2003

  • 2. EXAMPLES- FRAME: 13Inheritance: Similar to OOP Paradigm.

    Hotel Room

    what room

    where hotel

    contains

    hotel chair

    hotel phone

    hotel bed

    Hotel Chair

    what chair

    height 20-40cm

    legs 4Hotel Phone

    what phone

    billing guest

    Hotel Bed

    what bed

    size king

    part mattressMattress

    price 100$

  • 3. SLOTS:o Slots Denote Attributes.

    o Attributes are typed.

    o Some Values: Already Instantiated.

    o Some Values: Defined in the Instances only.

    o Instantiated Slots: May be Used to AnswerQueries.

    14

  • 3. SLOTS-II:Slots Can Contain:

    o Values

    o Types

    o Constraints Over Possible Values

    o Predicates:

    o Daemons/ Functions.

    o Slots Can be Structured- Frames Within the Slots-Frame Pointers.

    15

  • 3. SLOTS-III:o Constraints over Possible Values:

    BOY:

    16

    Is_a Human_being Type Constraints

    Name String

    Age Float

  • 3. SLOTS-IV:Predicates in a slot:

    o Two Types of Predicates in General:

    o If needed

    o If added.

    o If needed: State How to Find a Value.

    o If added: Used to check constraints on the Values.

    17

  • 3. SLOTS-IV:o Predicates in a slot- Example:

    BOY:

    If needed compute CGPA(BOY, results)

    If added test(constraint).

    18

    Is_a Human_being Type

    Name String

    Age Float

  • 4. INHERITENCE: 19Question: Does Tom drink Milk?

    TOM is an instance of BOY.

    Also, Boy is a HUMAN BEING.

    HUMAN BEINGS drink milk.

    All Human Beings Drink Milk.

    The answer can be inherited.

  • 5. GENERIC and DEFAULT VALUES:o Generic Values:

    o Hold True for all instances and sub-classes of the class.

    o Inherited Down the Hierarchy

    o Default Values:

    o Filler Values

    o Default Values may be over-written.

    o Specified in the Parent-Class(Super Class).

    o Conflicts: Multiple Inheritance.

    20

  • 6. INFERENCE IN A FRAME:o Look for Candidate Frame or Instances.

    o Look for Availability of Values.

    o If Available- Solved.

    o Else go to Preceding Level and so on.

    o If Solved at a Higher Level, try to have MoreSpecific Answers Going Down the hierarchy.

    21

  • 7. ANSWERING QUERIES- EXAMPLE: 22

  • 9. ADVANTAGES: 23o Makes programming easier by grouping related

    knowledge

    o Easily understood by non-developers

    o Expressive power

    o Easy to set up slots for new properties and relations

    o Easy to include default information and detect missing values.

  • 9. DISADVANTAGES: 24o No standards (slot-filler values)

    o More of a general methodology than a specific representation:o Frame for a class-room will be different for a professor and

    for a maintenance worker

    o No associated reasoning/inference mechanisms

  • 10. REFERENCES: 251. Artificial Intelligence, Lecture-20, Prof. Dr.

    Anupam Basu, Department of Computer Scienceand Engineering, IIT Kharagpur, NPTEL, MHRD,Government of India

  • 11. QUERIES?? 26

  • THANK-YOU