Mostakim, MoinWasi, Ahamed AlFahim, Ezazul HaqueInova, Nishat TasnimFahim, Abdullah AlPreeti, Taghrid Tahani2024-05-202024-05-20©20242024ID: 21301745ID: 17101266ID: 18101112ID: 19101567ID: 19301189http://hdl.handle.net/10361/22878Cataloged from PDF version of thesis.Includes bibliographical references (pages 53-54).This thesis is submitted in partial fulfillment of the requirements for the degree of Bachelor of Science in Computer Science, 2024.This research introduces a hybrid recommendation system through sentiment analysis for Bangla long textual sentences. Social media, as a vast source of opinions, can be harnessed through sentiment analysis using deep learning techniques, overcoming language barriers and improving recommendation systems. The paper addresses challenges in Bangla sentiment analysis, such as the scarcity of datasets and linguistic nuances, proposing a model that combines LSTM, Bi-LSTM, and CNN for optimized text sequence classification. The study explores six neural network models (ANN, CNN, LSTM,Bi-LSTM,BERT,RCNN) overcoming obstacles in dataset quality and distribution. Challenges in data collection, model selection, and computational resources are discussed. The paper concludes with the acknowledgment of the evolving frontier of sentiment analysis in Bangla text, emphasizing the transformative potential with continued efforts to expand datasets and refine algorithms.66 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.CNNEmbeddingLSTMBiLSTMVectorizationTokenizationBERTNeural networks (Computer science)Deep learning (Machine learning)Thesis