nlp.txt

learn intent->action to start learning a particualr topic
	can say lets learn "X"
	->fullfillment handles the web dev part where we fetch the questions/phrases from
	->ideally this should be simply a bunch of questions with prefetched asnwers
	->system says question->answer and allows for user queries in between

callback intent->action to revise learning 
(user input parsing and understanding/cross ref with given are ripe opportunities for NLP)
	can say lets revise "Y"
	->fullfillment retrieves user based session data
	->fetches a subset of the questions and the times etc for specific user
	->prompt here involves system asking a question
		->user answering
			->system responding on the level of response(one of the many classifiers)
			->system also providing an option to tell the answer so user can know what this is about
				->system decides a score on user input and appropriately adjusts it in the db


quickQuiz->similar to callback but with q questions on t topic both specified by the user ->ideal for slot filling
	can say hey lets do a quiz on t with q questions
	->slotfilling till we get both the pieces of information
	->flow for the quiz is similar to callback
		->Ambitious(can average user speaking speed and time the answers with some buffer)
		(so user can also specify time for each session, ie lets have a 5 minute revision session/learning session/quiz session)