[Next] [Up] [Previous]

Exercises

1. Figure 6.1 showed a very simple taxonomic tree for animals. Using a dictionary or encyclopedia if needed, try extending the tree to include `platypus', `iguana', and `zebra'. What special problem is posed by trying to create a node for `platypus' in the hierarchy? How might this be overcome? (Hint: Look again at the treatment of `ostrich' in this chapter.)
2. In the second section of this chapter's appendix, we have shown how semantic networks can be represented as lists, and we have represented facts about Tweetie the Canary in this form. Complete the list representation of the ISA-hierarchy in figure 6.1 with facts about Percy the Platypus, Iggy the Iguana, and Ziggy the Zebra.
3. On page [*] we gave examples of syllogisms, and in figure 6.4 we showed how the set-theoretic interpretation of birds and canaries could be represented diagrammatically with circles. We might regard such a diagrammatic representation as a `model' of facts given in syllogistic form. Draw similar diagrams to prove that the conclusions in the syllogisms below do not logically follow from the premises.

   1.  
      Premise 1:    		 All who eat cream cakes get fat.
      
      Premise 2:    		 Billy doesn't eat cream cakes.
      
                    		 0.4em
      
      Conclusion:   		 Billy isn't fat.
      
      

   2.  
      Premise 1:    		 Anyone who studies AI is clever.
      
      Premise 2:    		 Molly is clever.
      
                    		 0.4em
      
      Conclusion:   		 Molly studies AI.
      
      

4. Using the definition of answer in the appendix to this chapter, as a first model, try to write a procedure that, given your database of facts about animals, will answer questions such as

   Is a shark a fish?
   Does a zebra have gills?
   Can an iguana fly?

5. Suppose an impecunious tourist at Victoria wishes to know which tube stations can be reached for just 60 pence. You might imagine that one way for the Automated Tourist Guide to find out would be to use spresent as before, but this time inserting a variable in the place of the name of the station. For example, instead of

   spresent([[brixton] fare 60]) =>

   you might feel you could ask

   spresent([?station fare 60]) =>

   However, when you then type in

   station =>

   you may be surprised by the answer:

   ** [zone 2]

   Explain in detail why you get this response.

[Next] [Up] [Previous]