Comparative Analysis for Text Summarization of Educational Websites Content Using Pretrained BART Model

Educational websites offer diverse learning materials, but searching and summarizing them can be time-consuming and challenging. Unrestricted access to large websites poses risks for younger users. Traditional summarization methods like TF-IDF and BART with TextRank require significant memory and processing time, making them inefficient for large datasets. This study introduces a pretrained BART model for efficient text summarization of educational website content. The proposed model leverages the power of a large language model which is pretrained on a massive dataset, improving efficiency and accuracy. The model focuses on extracting key information from educational websites while ensuring factual correctness and conciseness. The performance of the model is evaluated using standard summarization like BLEU score and ROUGE metrics. The pretrained BART model outperforms traditional methods on educational website texts in terms of both computational and efficiency.