摘要：Using sentiment analysis, opinion mining examines the emotional tone and polarity of text (positive, neutral, or negative), as well as the sentiment polarity of text. With the rise of online information, sentiment analysis of customer evaluations has become a hot topic among machine learning researchers. Review texts for products online express a wide range of feelings and thoughts. By using natural language processing tools on the Internet, it will be possible for natural language processors to extract useful information from online reviews by performing sentiment analysis. It assigns polarity to a positive or negative entity or item. From the product reviews collected on Amazon, we conduct a Sentiment Analysis. As a result of asymmetrical weighting, we feed our feature words to support vector machine classifiers as well as Recurrent Neural Networks-Long Short Term Memory (RNN-LSTM)-optimised methods for determining the sentiment direction of reviews.
关键词：Sentiment Analysis;Natural Language Processing;Support Vector Machine;Recurrent Neural Network - Long Short-Term Memory
