Using justification patterns to advise novice UNIX users

Novice unix users have many incorrect beliefs about unix commands. An intel ligent advisory system for unix should provide explanatory responses that c orrect these mistaken beliefs. To do so, the system must be able to underst and how the user is justifying these beliefs, and it must be able to provid e justifications for its own beliefs. These tasks not only require knowledg e about specific unix-related plans but also abstract knowledge about how b eliefs can be justified. This paper shows how this knowledge can be represe nted and sketches how it can be used to form justifications for advisor bel iefs and to understand justifications given for user beliefs. Knowledge abo ut belief justification is captured by justification patterns, domain-indep endent knowledge structures that are similar to the abstract knowledge stru ctures used to understand the point behind a story. These justification pat terns allow the advisor to understand and formulate novel belief justificat ions, giving the advisor the ability to recognize and respond to novel misc onceptions.