Enhancing Content Recommendation System using Semantic Awareness with Flamingo Search Optimization and Bi-LSTM

A new approach for recommending appropriate content to e-learners. This system is designed to analyze various features of e-learners and determine the best possible content for them to learn. To accomplish this task, the Semantic similarity score aware Flamingo Search optimization Algorithm (SFSOA) is used. This algorithm is particularly useful in identifying the most relevant features that can be used to predict the most suitable content for the e-learners. In addition to the SFSOA, the proposed system also utilizes an ensemble deep classifier. This classifier is comprised of a Bidirectional Long Short-Term Memory (Bi-LSTM) and a fuzzy rule-based classifier. The Bi-LSTM is a deep neural network that can process sequential data and is particularly useful for tasks that involve natural language processing. The fuzzy rule-based classifier, on the other hand, uses a set of rules to make decisions based on uncertain or incomplete information. The proposed system was evaluated through experiments, and the results showed that it outperformed existing content recommendation systems in terms of accuracy of 99% than the previous recommendation has range from 92% to 97%. To conduct the experiments, a variety of question prompts were utilized along with the Amazon dataset, which contains evaluations of various topics. Five separate experiments were conducted on this dataset to investigate different sets of relevant documents for user inquiries. The experiments also involved varying numbers of documents, ranging from 100 to 500, which included both positive and negative content. In order to ensure a comprehensive analysis, the experiments covered content levels ranging from basic to advanced concepts Overall, this study presents a new and effective approach for content recommendation in e-learning, which can ultimately lead to better learning outcomes for students. The system is designed to identify the most relevant and influential features for predicting suitable content. An ensemble deep classifier, combining Bi-LSTM and a fuzzy rule-based classifier, is employed to make the final decision about the recommended content.