Networks-Based Representation

Abstract

Network-based method is another approach for knowledge representation and reasoning. They have particularly the advantage that, using the network one can navigate through the knowledge represented, and can perform the inferences. This chapter presents the semantic networks, conceptual graphs, frames, and conceptual dependencies, as well as their syntax and semantics. The $\mathscr {DL}$ (description logic)—a modified predicate logic for real-world applications is treated in detail, with examples of its language—the concept language for inferencing. Conceptual dependency (CD) is a language-independent representation and reasoning framework, such that whatever may be the natural language used, as long as its meaning is the same, the CD will be the same. The script language for representation and reasoning along with its syntax, semantics, and reasoning for CD is presented, followed with chapter summary, and an exhaustive list of exercises.

Exercises

• 1. Explain the difference between Ontologies and Semantic networks.

• 2. Describe the logical, structural, semantic, and procedural parts of semantic networks.

• 3. There are many words in the English language which can be used as noun and verb, for example, “book” in “Book my ticket” and “This is my book” have used the word “book” as verb and noun, respectively. In the following words, what are their different parts of speech?

  milk, house, liquid, airborne, group, set.

  Suggest a method in each case, as to how you will reason the true meaning.

• 4. Suggest a data structure for the implementation of semantic nets such that retrieval can be as fast as possible.

• 5. Represent the relationships between quadrangle, parallelogram, rhombus, rectangle, and square in the form of a semantic network. Is the semantic network unique, or are there many different forms it can take?

• 6. Represent the following statements using semantic networks:

  • a. “Rajan teaches his students a lot of innovative things.”

  • b. “Raman tells Rajan’s students a number of useful things.”

  • c. Mike and Mary’s telephone number is the same.

  • d. John believes that Mike and Mary’s telephone number is the same.

• 7. Represent the following knowledge in a semantic network:

  Dogs are Mammals	Birds have Wings
  Mammals are Animals	Bats have Wings
  Birds are Animals	Bats are Mammals
  Fish are Animals	Dogs chase Cats
  Worms are Animals	Cats eat Fish
  Cats are Mammals	Birds eat Worms
  Cats have Fur	Fish eat Worms

• 8. Represent the following in partitioned semantic networks:

  • a. Every player kicked a ball.

  • b. All players like the referee.

  • c. Andrew believes that there is a fish with lungs.

• 9. Represent the following statements using semantic networks:

  • a. “John tells his students a lot of useful things.”

  • b. “Andrea tells John’s students an enormous number of useful things.”

  Suppose you wanted to build an AI system that was able to work out “who tells John’s students the greatest number of useful things.” How could you do that?

• 10. Suppose you learn that “Tom is a cat”. What additional knowledge about Tom can be derived from your representation? Explain how.

• 11. Suppose Tom is unlike most cats and does not eat fish. How could one deal with this in the semantic network?

• 12. Formulate the solutions as to how the semantic networks can be used in the following cases?

  • a. Natural language understanding

  • b. Information retrieval

  • c. Natural language translation

  • d. Learning systems

  • e. Computer vision

  • f. Speech generation system

• 13. “The inferencing in semantic networks make use of unification, chaining, modus ponens, and resolution.” Justify each, taking a suitable example.

• 14. Explain, using semantic networks, how we can map an object’s perception to concepts, and identify these concepts. Give examples.

• 15. How semantic networks help in understanding the meaning of words in natural language sentences? Explain.

• 16. Represent the following as a series of frames:

  Dogs are Mammals	Birds have Wings
  Mammals are Animals	Bats have Wings
  Birds are Animals	Bats are Mammals
  Fish are Animals	Dogs chase Cats
  Worms are Animals	Cats eat Fish
  Cats are Mammals	Birds eat Worms
  Cats have Fur	Fish eat Worms

• 17. Express the following sentences in Description Logic:

  • a. All employees are humans.

  • b. A mother is a female who has a child.

  • c. A parent is a mother or a father.

  • d. A grandmother is a mother who has a child who is a parent.

  • e. Only humans have children that are humans.

• 18. Translate the logic expressed in Fig. 7.10 into $\mathscr {DL}$.

• 19. Select one or more answers from the following. Also, justify the answer(s) selected by you.

  • a. What type of reasoning is performed using semantic networks?

    (A) Deductive	(B) Default
    (C) Inductive	(D) Abductive
    (E) Hierarchical

  • b. In the Description Logic, the domain is always,

    (A) Open world	(B) Closed world
    (C) Depends on the domain used	(D) None of above