Intelligent Systems - Review Questions 2

This is the second set of review questions. As with earlier questions they are meant to provoke a bit of thought, so dont just give one word answers.


Knowledge Representation (KnRpn)

  1. which KnRpn systems are most appropriate for the following (think about the representational form and the necessary inference mechanisms)
  2. identify 3 cases where human sensory perception effects an emotional response. Consider the types of Kn/memories involved in each case and consider the implications for computerisation
  3. human cognition is bounded by limitations in attention, memory types, familiarity & field dependence. What does all this mean? Can these issues affect computer based systems?
  4. think of suitable representations for each of the following (try to think of more than one in each case)
  5. provide an example of each of the following and state what they are most suitable for
  6. how does a semantic network handle default values?
  7. what is a demon? how do demons interact with inheritance?
  8. what purpose is served by perspectives? how can they be dealt with in practice?
  9. what is the purpose of generalisation? How is it achieved?
  10. what problems can occur with generalisation? How can these be minimised?
  11. how can semantic nets be compared for similarities? Differences?
  12. what kind of processing underpins the following tasks


Language Processing

  1. give 3 examples each of syntactic, semantic & pragmatic ambiguity

  2. discuss the problems associated with each of the following sets of sentences and comment on how the problems can be dealt with
  3. how did the early, pattern matching language processors work? why dont we still use that approach?
  4. what are the objectives of morphological processing? why bother?
  5. identify 3 types of language feature which influence morphological processing
  6. what is meant by case?
  7. what is the purpose of syntax analysis and what tools are used in this process?
  8. identify 5 different syntactic word categories and give an example word for each one
  9. identify 3 different phrase types and give an example of each

  10. write the grammar rules and lexicon for the following sentences, draw the phrase structure (parse-tree) for each one

  11. discuss the differences in the following pairs of sentence, how is this dealt with (i) in language processing (ii) with Lkit?
  12. go over the tutorial exercises based on Lkit - you should be able to write lecon entries and grammar rules
  13. how are semantics expressed using Lkit?
  14. how is it possible to derive actor-action-object case frames from simple sentences?
  15. how do fails & glitches work?
  16. how can you use Lkit to recognise passive as well as active sentence types?
  17. how are complete case frames produced from raw sentence case frames?
  18. what are the following cases used to describe: source, start-time, instrument, beneficiary?
  19. name 3 other cases
  20. write two alternative primitive verb forms for "kick"
  21. describe the types of inference that occurs after a verb form has been disambiguated
  22. describe dialogs and event sequences