Question and answer dialogue generation for intelligent tutors
OwnerBenny G. Johnson, Dale A. Holderinventors
Life of this patent
Estimating the term…
Abstract
A question generation means is developed for intelligent tutors wherein a question menu having dynamic questions is capable of being displayed after each step of a problem. The generated question set contains both elementary and advanced questions and changes after each step is performed, thereby enhancing the potential that a student using the tutorial understands every aspect of a problem.
Claims6
A computerized method for generating dialogue in an intelligent tutor, said method comprising:
- accepting an externally supplied problem from a student, wherein said problem is the student's own problem;
- receiving a step towards a solution to said problem as input;
- creating a question-generating rule class as a derived class for generating questions for selection by said student as a form of said input;
- providing a meaningful response to said input even when said input towards said solution to said problem is unanticipated, further comprising the step of:
- interfacing each of said question-generating rule, production rules to model anticipated errors and consistency rules to impose constraints on a solution to said problem;
- testing the applicability of said production rule, said consistency rule, and said question generating rule after each said input is received;
- and
- generating said question relevant to the state of said tutor from both context-specific and pedagogical factors to augment and enhance the actions of said production rule and said consistency rule such that different paths of inquiry are made possible for the same problem.
A computer-readable medium having computer executable instructions for performing a method for generating dialogue in an intelligent tutor, said method comprising:
- accepting an externally supplied problem from a student, wherein said problem is the student's own problem;
- receiving a step towards a solution to said problem as input;
- creating a question-generating rule class as a derived class for generating questions for selection by said student as a form of said input;
- providing a meaningful response to said input even when said input towards said solution to said problem is unanticipated, further comprising the step of:
- interfacing each of said question-generating rule, production rules to model anticipated errors and consistency rules to impose constraints on a solution to said problem;
- testing the applicability of said production rule, said consistency rule, and said question generating rule after each said input is received;
- and
- generating said question relevant to the state of said tutor from both context-specific and pedagogical factors to augment and enhance the actions of said production rule and said consistency rule such that different paths of inquiry are made possible for the same problem.
Dependencies are inferred from the claim text.