Closed Brtelfer closed 2 years ago
The problem is that the system allows for single letter edit distances (adding or omitting a letter from the answer). Because of the lengths of these answers it will assume you were close enough as long as one letter is correct. We can check for a minimum length of an answer to allow or deny the use of single letter edit distances.
Exactly, the originally algorithm for lf parameter didn't do that unless it was always broken. It is not behaving as I designed it
On Wed, Jun 22, 2022 at 1:13 PM Rhodes White @.***> wrote:
The problem is that the system allows for single letter edit distances (adding or omitting a letter from the answer). Because of the lengths of these answers it will assume you were close enough as long as one letter is correct. We can check for a minimum length of an answer to allow or deny the use of single letter edit distances.
— Reply to this email directly, view it on GitHub https://github.com/memphis-iis/mofacts-ies/issues/841#issuecomment-1163454851, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADDLPK6QQ7CTFYX2SIXOVETVQNJU5ANCNFSM5ZOIRNDA . You are receiving this because you were assigned.Message ID: @.***>
-- Philip I. Pavlik Jr. @.*** http://optimallearning.org/
OK, I looked it over, and lets just remove this code "|| editDistance <= 1"
Not sure why that was ever wanted.
Ifparameter is not working properly for the Hiragana and Katakana test on staging. When is set to .85, it acts like it is set to .5. For instance, when the answer is "ka”, it says “ha” is close enough or that “a” is close enough when the correct answer is “sa”. “h” was accepted when the right answer was “a”.