This project analyzes 3800+ British Airways customer reviews using Natural Language Processing (NLP), Sentiment Analysis, and Topic Modeling to uncover key trends in customer sentiment. The goal is to provide data-driven insights to improve customer satisfaction by identifying major pain points and areas of excellence.
✔ Perform Exploratory Data Analysis (EDA) to uncover patterns in customer reviews.
✔ Apply Sentiment Analysis to classify reviews as Positive, Neutral, or Negative.
✔ Use Topic Modeling (LDA) to extract key themes from unstructured text.
✔ Develop data-driven recommendations to enhance customer experience.
- Source: Scraped customer reviews from Skytrax
- Size: ~3,900+ reviews
- Challenges: Unstructured text, missing values, and redundant data
- Programming Language: Python 🐍
- Libraries: Pandas, NumPy, Seaborn, Matplotlib, Scikit-learn, NLTK, TextBlob, WordCloud
- Techniques: NLP, Sentiment Analysis, LDA Topic Modeling, Data Visualization
- 67% positive ratings, with praise for crew friendliness, comfort, and food quality.
- 32% negative ratings, highlighting issues like delays, cancellations, and poor customer service.
- 1% neutral, indicating mixed experiences.
1️⃣ Positive Experiences – Crew service, food, and business class comfort.
2️⃣ Customer Complaints – Cancellations, refund issues, and poor customer service.
3️⃣ Flight Experience – Long wait times, security checks, and boarding delays.
4️⃣ Seating & Class Experience – Complaints about economy seating discomfort.
5️⃣ Lounge & Airport Experience – Heathrow Terminal 5 feedback, lounge quality, and wait times.
How British Airways can improve based on data insights:
✔ Improve Customer Support Response Time – Many complaints about slow refunds & issue resolution.
✔ Enhance Communication on Flight Status – Frequent frustration with unexpected cancellations.
✔ Increase Comfort in Economy Class – Seat comfort was a recurring complaint.
✔ Optimize Operations & Baggage Handling – Customers expressed dissatisfaction with lost luggage.
- Strengthen brand loyalty & increase customer satisfaction
- Enhance customer satisfaction
- Fewer negative reviews & social media complaints
- Enhanced brand reputation & customer retention
💡 Build a Tableau or Power BI executive dashboard that lets users:
- Filter sentiment by route,
- Compare review themes over time,
- Drill into negative reviews by topic,
- Monitor sentiment trends,
- Use Deep Learning (BERT, GPT) for more accurate sentiment classification.
- Implement automated alerts for sudden spikes in negative reviews.