Opinion mining, also known as sentiment analysis, plays a crucial role in understanding and analyzing public sentiment towards various products, services, or topics. This paper proposes a hybrid approach that combines deep learning techniques with machine learning algorithms to perform effective opinion mining. The first step in the approach is dataset preprocessing, where stemming and lemmatization are applied using natural language processing (NLP) techniques. Stemming and lemmatization help reduce word variations and normalize the text data, enabling more accurate analysis.