Quality of Questions #4
Description
There seems to be an abundance of bad/obvious questions.
Questions don’t seem to follow easy-med-hard guidelines.
User-rated-question-feature needs to feed the prompt after submission in order to train the model the differences between good and bad before a final implementation.
Since patterns are emerging after manually using the app, it may indicate a ‘categorization’ of ‘types’ of good and bad questions which could be used to train.
User inputs topic (user) -> questions are generated (prompt) -> user answers questions (user) -> user rates questions (user) -> user submits feedback (prompt)