TITLE : Sentiment Analysis for Customer

Reviews on Amazon Customer Service

(2024) Using AI.
BY G N VISHWANATHA GOWDA, KUSUMA R, KEERTHANA P, HEMANTH H N, HEMANTH GOWDA GK
Introduction to Sentiment Analysis Sentiment Distribution by Month
MBA, B SECTION, SAPTHAGIRI NPS UNIVERSITY, BENGALURU Key Trends in Customer Feedback
Analysis of the textual content within customer reviews reveals consistent patterns in both positive
Sentiment analysis, also known as opinion mining, is a natural language processing (NLP) technique used to determine the  Highest Positive Ratings: The month with the highest and negative feedback. These recurring themes provide valuable insights into specific aspects of
emotional tone behind a piece of text. It helps identify sentiments such as positive, negative, or neutral in customer reviews, positive ratings is October (1,369). Amazon's customer service that are either exceeding expectations or requiring improvement.
social media posts, or other textual data. Businesses use sentiment analysis to understand customer opinions, enhance  Lowest Positive Ratings: The month with the lowest Common Positive Feedback Trends
marketing strategies, and improve customer service. The process involves machine learning, text analysis, and computational positive ratings is November (1,104). 1."Amazing Experience" - Customers frequently praise exceptional
linguistics. Sentiment analysis is widely used in fields like finance, healthcare, and social media monitoring.  Highest Negative Ratings: The month with the highest service interactions that exceed expectations.
negative ratings is May (461). 2. "Excellent Support" - Technical assistance and problem-solving
 Lowest Negative Ratings: The month with the lowest capabilities receive consistent positive mentions
3."Great Service!" - General satisfaction with the overall service
Study Objectives negative ratings is November (378).
 Over all ratings show a slight decline in November
quality and responsiveness appears regularly
This sentiment analysis project aims to provide a comprehensive (720), which is the lowest for the year. Common Negative Feedback Trends
understanding of customer perceptions regarding Amazon's customer 1.”Very Disappointed" - Unmet expectations and service failures
service. By analyzing a large dataset of customer reviews, we can generate strong negative reactions
identify patterns, trends, and areas for improvement that might Overall Sentiment Distribution 2."Slow Response" - Delays in addressing customer inquiries or
otherwise remain hidden in the vast amount of textual feedback. The analysis of 10,000 customer reviews reveals a relatively resolving issues are frequently cited
balanced distribution of sentiment across the three main 3."Unhelpful" - Customers report interactions that failed to
categories. This even distribution provides valuable insights address their specific needs or concerns
into the overall customer perception of Amazon's customer
service performance in 2024.
The distribution of customer sentiment shows a remarkably
Insights & Recommendations
balanced pattern across all three categories. Positive reviews
Based on the comprehensive analysis of 10,000 customer reviews, several key insights
Dataset Overview. account for 33% of the total (approximately 3,300 reviews),
while negative reviews also represent 33% (around 3,300 emerge that can guide strategic improvements to Amazon's customer service operations.
This analysis is based on a substantial dataset of customer reviews specifically focused reviews). Neutral reviews form the slightly larger segment at
approximately 34% (around 3,400 reviews).
on Amazon’s customer service performance. Understanding the composition and
This even distribution suggests that while Amazon's customer
 The Address Neutral Experiences
source of this data is essential for contextualizing the findings and insights presented service succeeds in creating positive experiences for a third of  Investigate Negative Feedback Further
throughout this report. its customers, there remains significant room for  Improve Resolution Speed & Effectiveness
1.Company:amazon pvt ltd , improvement. The substantial proportion of neutral reviews
2.dataset size:10000, represents an opportunity
3.Sentiment Categories:Reviews are classified into three distinct sentiment categories:
Positive, Neutral, and Negative, allowing for clear segmentation and
Sentiment Trends Over Time
comparative analysis
4.Data Source:This dataset was provided by ChatGPT AI
Conclusion ->>>
for learning purposes only and should be considered temporary and illustrative rather
than actual Amazon customer data.
 The line chart visualization provides a Sentiment analysis provides a clear picture of customer experiences and identifies specific areas for
dynamic view of how customer sentiment service enhancement. By addressing the identified issues and improving response efficiency,
has evolved throughout 2024. This temporal Amazon can enhance customer satisfaction and strengthen brand loyalty. The balanced distribution
Methodology analysis reveals important patterns and of sentiment suggests
fluctuations that can help identify seasonal significant opportunity for improvement, while the consistent patterns in feedback provide clear
trends, the impact of specific events, or direction for targeted interventions.
changes in service quality over time. By converting neutral experiences to positive ones and addressing the root causes of negative
feedback,
Amazon can significantly improve overall customer perception and potentially gain competitive
1.Monthly Fluctuations: While sentiment shows some variation across different months, the overall advantage through superior customer service.
pattern remains relatively stable throughout the year. This suggests that Amazon's customer service
maintains a consistent level of performance despite seasonal changes and varying demand.
2.October Peak :October 2024 stands out with a significant increase in positive reviews (306 more than
the previous month). This peak could be attributed to successful service improvements, seasonal
staffing adjustments ahead of the holiday season, or the resolution of previous issues.
TEAM
3.November Dip: November experienced the lowest ratings across all sentiment categories. This
universal decline might indicate increased service pressure during the early holiday shopping season,
staffing challenges, or system issues affecting customer experience across the board
4.Negative Peaks: Negative sentiment reached notable peaks during August and October, suggesting KEERTHANA P HEMANTH H N HEMANTH GOWDA GK
45
G N VISHWANTHA GOWDA KUSUMA R
specific service issues or challenges during these periods.
MBA,SAPTHAGIRI NPS UNIVERSITY, BENGALURU