Let's assume that S is some sentence (e.g. theorem). Are there efforts to quantify how well some agent (pupil, student, machine agent) understands this sentence S (or some sent of sentences)? E.g., if student is capable to derive many other sentences that are conclusions of the S+preexisting_knowlege_base_in_the_head_of_the_student, then, obviously, her understanding of S ir higher. Hence - maybe we can quantify the understanding by the cardinality (volume) of the set of conclusions that can be derived from compounded knowledge base (and use the volume of the initial base as the normalizing factor/divisor).

Am I alone in my efforts to formalize and quantify the understanding and comprehension of statements or are there other efforts? What are those efforts? What ar common terms, name, scientists in this field?

I guess, that such notions can have application in education sciences, e.g. to devise frameworks and tools for boosting the comprehension.

  • $\begingroup$ So you wish to measure per-sentence comprehension on-the-fly or something? I mean, exams test whether you have mastered the material in the long run so that would be the obvious answer. If you talk about a per-sentence on-the-fly evaluation then it would almost become a speech test, like 'repeat the words I just told you'. I'm failing to see the practical framework what you wish to achieve here. $\endgroup$ – AliceD May 5 '19 at 18:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.