What it does
Flash Notes uses natural language processing API's to automatically pick out important terms in the students notes and generate related questions that can be used as a study guide.
How we built it
We used the Microsoft Azure API to pick out "key terms" from the original notes. We then sorted these key terms in order of appearance in the original notes and compared them to the "mood" output from the google natural language API. We compared both outputs to locate questions and answers related to said "key terms".
Challenges we ran into
It was quite difficult to design algorithms that would be able to create question-like phrases like 'fill in the blanks' or 'define this'
Accomplishments that we're proud of
Setting up the API's and harnessing their potential fairly well to our cause. We are also proud of the algorithms we developed to pick questions and their answers.
What we learned
We learned how to integrate API's into full stack application and furthered our algorithmic knowledge.
What's next for Flash Notes
We would love to generate more types of questions, for example diagrams, mathematics and/or matching, which would allow for the creation of extensive quizzes on the course material.
We were inspired by the lack of teacher specific course quizzes and flashcards available for students, often the available resources are not completely relevant to the content covered in class.