Aspect Based Sentiment Analysis for Amazon Data Products using PAM

Abstract

Recent years have seen a considerable evolution in online shopping, as well as the ability to get text-based client feedback, comments, and recommendations. The sentiment analysis method examines a large collection of text data to determine the customer’s opinion. One method for analyzing customer sentiment that is provided by natural language processing is AbSA. Understanding out precisely exactly consumers feel about a product is the aim of AbSA. In this research, we spoke about aspect-based sentiment analysis, which identifies the sentiment by using the aspect term. To illustrate our framework, we used user reviews for headphones and earphones that were gathered from the Amazon website. We used NLP to preprocess the data, and we then used the Pachinko Allocation model (PAM) to extract the aspect term and polarity from the review. Machine learning technique have been explored to evaluate the model.