Karim, Dewan ZiaulAhmed, Md FaisalHassan, MehediPritom, Mujtaba WasifFuad, Shahidul IslamSifat, Saif Ahmmed2025-02-202025-02-2020242024ID 20201148ID 20201130ID 20201055ID 21101341http://hdl.handle.net/10361/25480Cataloged from PDF version of thesis.Includes bibliographical references (pages 54-55).This thesis is submitted in partial fulfillment of the requirements for the degree of Bachelor of Science in Computer Science and Engineering, 2024."These days, customers are more keen to buy products online rather than going to a shop or market. However, they often fear about the quality of products, as there is no way to measure them before buying them. As a result, most buyers rely on the reviews of other customers who have already purchased the product. For this reason, customer reviews are very crucial for the e-commerce industry. A popular method for assessing the quality of a product is opinion mining, which is also called sentiment analysis. It is a method of extracting emotion from a text using natural language processing (NLP). In this study, we have collected data from an e-commerce site named Daraz and introduced a new dataset that contains Bengali reviews. A total of 48000 reviews were collected, of which 22000 were Bengali. 15000 are in English, while the rest of 9000 are in “Banglish” (Romanized Bengali). Several data preprocessing techniques were used to introduce a new clean dataset that only contains Bengali reviews. Five machine learning algorithms—Naive Bayes, Random Forest, Gradient Boosting Classifier, Logistic Regression, and Support Vector Machine (SVM)—and three deep learning models—BiLSTM, Multilingual BERT, and BanglaBERT—were implemented to evaluate our work. Our work should help the sellers filter out the best products that are popular among consumers. "64 pagesenBRAC University theses are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission.NLPNaive BayesRandom forestGradient boosting classifierLogistic regressionSupport vector machineE-commerceThesis