Recommender systems enhance the user experience by integrating factual and behavioral practices, enhancing learning strategies, and granting academic opportunities based on learners subjective interests. Cognitive intelligence, a learner’s capability to interpret and synthesize cognitive abilities, is crucial to assemble effective recommendations. However, a student encounters diverse adversities in his or her academic life, whether it is his or her academic pacers or outward circumstances. As a result, a learner consequently experienced severe academic setbacks. Thus, this paper proposes a novel approach to identify cognitive deficits of learners in a school or college before recommending any scholastic content or pertaining in any academic activity. We have employed a modified Bi-directional Gated Recurrent Unit (Bi-GRU) method in conjunction with Shuffled Frog Leaping Algorithm (SFLA). Our approach overcomes some of the shortcomings of conventional Bi-GRU method in choosing the right scaling parameters during training. Bi-GRU method is opted for detecting the cognitive deficits of a learner and SFLA is applied over Bi-GRU in order to hyper-Tune the parameters. The proposed method is compared with a few cutting-edge methods, and the experimental result outperforms the other methods in training, validation, and testing accuracies, achieving 99.34%, 98.98%, and 99.19% respectively.