A Deep Learning-Based Sentiment Analysis Framework for Online Product Ranking Using Probabilistic Linguistic Term Sets
Main Article Content
Abstract
This manuscript Formulate an innovative methodology employing deep learning and sentiment analysis to improve online product rankings, tackling the issue of excessive e-commerce options and offering a more user-focused assessment that incorporates subjective satisfaction elements. Employ advanced deep mastering methodologies to increase a resilient sentiment evaluation model educated on a diffusion of user evaluations. Incorporate sentiment scores into the current product ranking algorithm, creating a hybrid methodology that merges objective metrics with subjective user experiences. Utilise sentiment analysis to extract insights from unstructured data, providing customers with detailed product assessments and delivering meaningful feedback to businesses. Connect user comments with rankings to facilitate more informed and satisfactory purchasing selections. Anticipate an enhanced, user-centric on-line product ranking system providing tailored recommendations. Deploy various deep learning models and evaluate their efficacy using established measures, thereby benefiting primary stakeholders including ecommerce platforms, businesses dependent on user input, and consumers in pursuit of educated product recommendations. We enhanced the project using a highly precise LSTM “Hybrid model (LSTM + GRU), attaining 99.9% in accuracy, precision, recall, and F1-score”. to improve user testing, we created an intuitive Flask-based front end with authentication capabilities, guaranteeing a secure and tailored user experience.