✨ Amazon Product Review Analysis: Business Intelligence Solution ✨

🚀 Project Overview

This repository contains a comprehensive Business Intelligence (BI) solution focused on analyzing Amazon product reviews, developed as part of the PUSL2079: Topics in Business Analytics & Intelligence coursework at the University of Plymouth.

The project includes a detailed case study on Amazon's use of business analytics and a custom BI solution to analyze customer product reviews, delivering actionable insights to enhance customer satisfaction and product quality.

---

🎯 Objectives

• Conduct a case study on how Amazon leverages business analytics to address operational and customer-centric challenges.
• Design and implement a robust BI solution to identify patterns in customer reviews, focusing on negative feedback to improve product listings and reduce return rates.

---

🛠️ Project Scope & Implementation

The project was meticulously executed, covering every phase from problem definition to actionable recommendations. Key tasks included:

• 🎯 Problem Definition
• 📊 Data Collection & Selection (Amazon Product Reviews from Kaggle)
• 🧹 Data Cleaning & Preprocessing (Power BI Power Query + Python)
• 🧠 Sentiment Analysis (Python, TextBlob)
• 📈 EDA & Visualization (Power BI)
• 💡 Recommendations
• ✍️ Documentation & Referencing
• 🤝 Ethical Considerations

---

📂 Power BI Dashboard File

You can download and explore the Power BI dashboard file here:
Amazon Product Review Analysis - Power BI File

---

🧠 Skills Learned

As part of the Amazon Product Review Analysis: Business Intelligence Solution project, the following skills and tools were applied and developed:

🔧 Tools & Technologies

• Power BI
  • Dashboard Design
  • Power Query (Data Cleaning & Transformation)
  • DAX (Calculated Columns & Measures)
• Python
  • Data Preprocessing (pandas, numpy)
  • Sentiment Analysis (TextBlob)
  • Word Cloud Generation (matplotlib, wordcloud)

📊 Data Analysis & BI Skills

• Data Visualization
• Exploratory Data Analysis (EDA)
• Natural Language Processing (NLP)
• Business Intelligence Reporting
• Interactive Q&A Interface Development

📈 Business & Communication Skills

• Business Problem Definition
• Case Study Research
• Insight Generation & Interpretation
• Recommendation Writing
• Ethical Considerations in Data Usage
• Stakeholder-Oriented Reporting

---

📈 Exploratory Data Analysis & Visualizations (Power BI)

1️⃣ Histogram of Ratings (Figure 1)

Insight: Most reviews are 4-5 stars (1.82M for 5 stars, 0.32M for 4 stars), but 1-2 star reviews highlight improvement areas.
Action: Investigate low-rated reviews.

---

2️⃣ Pie Chart of Sentiment (Figure 2)

Insight:

• 76% Positive (432.03K)
• 19.28% Neutral (109.59K)
• 4.72% Negative (26.84K)
  Action: Leverage positive sentiment; fix negative patterns.

---

3️⃣ Bar Graph – Negative Reviews by Product (Figure 3)

Insight: Products like B000X1Q1G8 receive high negative reviews.
Action: Prioritize those for quality checks.

---

4️⃣ Line Chart – Sentiment Trend Over Time (Figure 4)

Insight: Positive sentiment fluctuated; others remained stable.
Action: Prepare for seasonal patterns.

---

5️⃣ Word Cloud – Negative Keywords (Figure 5)

Insight: Keywords like “broken,” “cheap,” and “return” indicate common issues.
Action: Improve product and delivery quality.

---

6️⃣ Bar Chart – Average Rating by Sentiment (Figure 6)

Insight:

• Positive ≈ 4.5
• Neutral ≈ 3.5
• Negative < 2.5
  Action: Focus on converting neutral to positive.

---

7️⃣ Interactive Q&A Interface

Insight: Enables real-time dynamic querying.
Action: Support stakeholders with fast, flexible insight.

---

💡 Recommendations

• Improve Product Quality – Focus on products with high complaints (e.g., B000X1Q1G8).
• Enhance Holiday Operations – Plan for seasonal demand spikes.
• Leverage Positive Sentiment – Use in branding and campaigns.
• Streamline Returns – Improve logistics and return experience.
• Refine Product Descriptions – Avoid misleading terms.
• Engage Neutral Customers – Deliver upgrades to improve ratings.

---

🤝 Ethical Considerations

• Data Privacy – No PII included.
• Transparency – All steps are documented.
• Fair Use – Used for academic learning purposes.
• Bias Mitigation – Validated sentiment analysis across product groups.

---

📚 References

• Amazon Web Services – Case Studies
• AWS – Machine Learning at Amazon
• Kaggle – Consumer Reviews of Amazon Products
• Forbes – How Amazon Uses Big Data
• Microsoft – Power BI Community Blog
• TextBlob – Documentation
• Provost, F., & Fawcett, T. (2013). Data Science for Business. O’Reilly Media.

---