BigMemo – plataforma web de memorização com repetição espaçada e classificação automática de dificuldade por IA

Abstract

Long-term retention of information is one of the greatest challenges for those who study. Classic experiments in cognitive psychology showed that human memory decays exponentially within hours or days after the acquisition of new knowledge, in the absence of active review. Techniques such as spaced repetition and retrieval practice are effective responses to this problem, with strong empirical support, but applying them manually is unfeasible for anyone managing hundreds of items simultaneously. This work presents BigMemo, a web platform created to help students and autodidacts organize, review, and retain content systematically. The system implements the SuperMemo 2 algorithm with a three-level difficulty scale, generates flashcards automatically through large language models, and provides a progress statistics panel. Two features differentiate BigMemo from the other SRS platforms analyzed: automatic difficulty classification by AI, which semantically evaluates the user’s written answer and mitigates the metacognitive biases inherent to self-assessment; and the estimated retention index per deck, derived from the forgetting curve and updated in real time, which makes visible the percentage of knowledge the student still retains. Together, these features position BigMemo as a differentiated alternative among the SRS platforms analyzed.