Sentiment Analysis on E-Commerce Sites

 

Objective

To analyse the sentiments of people on various e-commerce sites to understand the people’s view or Sentiment Analysis on E-Commerce Sites. This will help the e-commerce sites to enhance their method.

Project Overview

This project concentrates on Twitter sentiment analysis since it is a better approximation of public sentiment as opposed to conventional internet articles and web blogs. The reason is that the amount of relevant data is much larger for the twitter, as compared to traditional blogging sites.

Sentiment analysis of public is important in any business. This could be done by analyzing overall public sentiment towards that product with respect to time and using tools for finding the public sentiment. This can also estimate how well the product is responding to the market by classifying tweets into the positive, negative and neutral. Using this information, the product efficiency can be enhanced.

Proposed System

The sentiment analysis of the user on an e-commerce site is proposed to provide the better understanding and people’s view of the particular product by displaying the classified user sentiment at the high accuracy rate. This would help to purchase the better product by people and also help the product developers to enhance their product.The proposed system architecture is shown in the figure.

List of Modules for Sentiment Analysis on E-Commerce Sites 

• Data Gathering
• Attributes Collection
• Statistics
• Result and Analysis
• Data Visualization

Module 1: Data Collection

The dataset for this project is collected from the twitter using R tool for e-Commerce site. Data is collected for top three e-commerce sites such as Flipkart, Amazon, and Snapdeal. TwitterAPI is used to extract the data from Twitter.

Module 2: List of Attributes

The collected dataset consists of following attributes.

• Row no
• Id
• Polarity_connection
• Subjectivity_connection
• Polarity
• Subjectivity
• Created-At
• From-User
• From-User-Id
• To-User
• To-User-Id
• Language
• Text

 Module 3: Statistics

Sentiment analysis of e-commerce sites will be used as a recommendation to the new users or the existing users. Also, sentiment analysis will help a company to boom their business and provide better quality to their customers. Here the twitter texts are classified into Positive, Negative and Neutral. Positive opinion words are used to express desired feelings while negative opinion words are used to express undesired feelings.

Module 4: Result and Analysis

Thus the data collection on various products statistical view on tweets such as classification of positive, negative and neutral tweets has been performed. The classification algorithm will be used to obtain the final outcome.

Module 5: Data Visualization

The classified data is analyzed, and the result is represented in the form of a bar chart, pie chart, and graph and word cloud. The graphical plot of the sentiment using Tableau tool makes it easy to evaluate performance e-commerce sites. The graphical plot of the sentiment using business intelligence tool makes it easy to evaluate the performance of classification algorithms.

 Software Requirements

• R Programming
• TwitterAPI
• Tableau

Hardware Requirements

• Hard Disk – 1 TB or Above
• RAM required – 8 GB or Above
• Processor – Core i3 or Above