Sentiment Analysis of Customer Reviews for Online Stores That Support Customer Buying Decisions

Sentiment Analysis of Customer Reviews for Online Stores That Support Customer Buying Decisions

Geetha Manoharan, Subhashini Durai, Gunaseelan Alex Rajesh, Sunitha Purushottam Ashtikar
Copyright: © 2023 |Pages: 9
DOI: 10.4018/978-1-6684-6519-6.ch015
OnDemand:
(Individual Chapters)
Available
$37.50
Current Special Offers
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

This research was carried out in order to conduct a sentiment analysis on customer reviews for an online store. It is a technique that makes use of textual contextual mining to identify and extract information that is subjective. This type of analysis aids a company in understanding the attitudes of their customers toward their brand, products, and services. When it comes to making evidence-based decisions, sentiment analysis is taken to the next level by using count-based metrics. The study examines the key aspects of the product that their customers are concerned about, as well as the reactions or intentions that these customers have toward their brand and product. The analysis is carried out using a machine learning approach, specifically a supervised learning approach. Sentiment analysis is carried out using the decision tree technique. The findings assist decision makers in understanding the attitudes of customers toward a brand, a product, or a service. This assists them in determining their future business strategy, which will help them increase their sales and profits.
Chapter Preview
Top

Literature Review

Social media, according to Alkhaldi et al. (2022), is an important tool for people to discuss and share events in their communities. This study uses three social media networks to test sentiment analysis (Instagram, Snapchat, and Twitter). By analysing social media posts about the first Saudi cruises, the findings of this study will help us better understand how people felt about them. 1200 cleaned samples were examined for this experiment. The researchers used a variety of machine learning algorithms to classify the data as positive or negative, including multilayer perceptron, naive bayes, random forest, support vector machine, and voting. The RF algorithm achieved a 100% accuracy rate with both test options and oversampled Snapchat data, which is the highest classification accuracy. The algorithms were put to the test on three different datasets. All of the algorithms worked well. As a result, the findings indicate that 80% of the attitudes were positive, while 20% were negative.

According to Verma (2022), the community's use of social media to discuss and share information about various events is critical. This study uses three social media sites to assess sentiment analysis of Saudi cruise opinions (Instagram, Snapchat, and Twitter). By analysing emotional responses on social media, this study will provide insight into how passengers and viewers felt during their first Saudi cruises. Following cleaning, over 1200 samples were collected for this experiment. Machine learning algorithms such as Multilayer Perceptron, naive Bayes, random forest, support vector machine, and voting were used to categorise the data. Using oversampled data from Snapchat, the RF algorithm achieved 100 percent classification accuracy, making it the most accurate. On three different datasets, the algorithms were tested. The performance of each algorithm was exemplary. The results indicate that eighty percent of the attitudes were positive and twenty percent were negative.

Complete Chapter List

Search this Book:
Reset