Sentiment Analysis for E-Commerce Products Using Natural Language Processing

Abstract

Sentiment analysis is one of the ways to evaluate the attitude of consumers towards products and services. E-commerce businesses have grown to a larger level in recent years. Customers' opinions and preferences are collected to analyze them further to boost online businesses. Collecting real-time structured and unstructured data and performing sentiment analysis on them are challenging and need to be addressed. We have used PySpark, and resilient distributed dataset (RDD) based sentiment analysis using Spark NLP to address scalability and availability issues in sentiment analysis on the e-commerce platform. We have also used FLASK-based Restful APIs and Scrapy for web scrapping to collect useful data from an e-commerce site. Our findings indicate that the proposed method of Natural Language Processing (NLP) for e-commerce products in real-time has enhanced efficiency in terms of scalability, availability, and faster data collection