🏏 IPL Match Commentary — Sentiment Analysis

NLP-based Sentiment Analysis on IPL Match Commentary using VADER & TextBlob
Personal Data Science Portfolio Project

---

📌 About the Project

This project applies Natural Language Processing (NLP) techniques to analyze the sentiment of IPL match commentary — classifying each commentary line as Positive, Negative, or Neutral. Using both VADER and TextBlob models, the project reveals emotional patterns across teams, overs, and match events.

---

🎯 Objectives

• Clean and preprocess real IPL commentary text data
• Classify commentary sentiment as Positive / Negative / Neutral
• Analyze sentiment trends by over, team, and match events
• Visualize insights with WordClouds and sentiment charts

---

🤖 NLP Models Used

Model	Type	Best For
VADER	Rule-based NLP	Social media & sports text — handles exclamations, caps
TextBlob	Lexicon-based NLP	Polarity (-1 to +1) & Subjectivity (0 to 1) scores

---

🔍 Analysis Performed

#	Analysis	Description
1	Text Cleaning	Lowercasing, special char removal, whitespace handling
2	VADER Sentiment	Compound score → Positive/Negative/Neutral classification
3	TextBlob Analysis	Polarity & Subjectivity scoring per commentary
4	Sentiment Distribution	Overall % breakdown across all commentary
5	Team-wise Sentiment	Which teams generate most positive commentary
6	Over-wise Trends	Sentiment patterns across different match phases
7	WordCloud	Most frequent words in Positive vs Negative commentary
8	Model Comparison	VADER vs TextBlob agreement analysis

---

📊 Visualizations Built

• 🥧 Pie Chart — Sentiment distribution (Positive/Negative/Neutral %)
• 📊 Bar Charts — Team-wise and over-wise sentiment breakdown
• ☁️ WordClouds — Positive words vs Negative words
• 📈 Line Charts — Sentiment trend across match overs
• 🔥 Heatmap — Sentiment correlation matrix

---

🛠️ Tech Stack

Technology	Purpose
Python 3.8+	Core programming language
Pandas	Data loading & manipulation
VADER Sentiment	Primary sentiment classifier
TextBlob	Secondary NLP analysis
WordCloud	Word frequency visualization
Matplotlib	Base visualization library
Seaborn	Statistical visualizations
re (regex)	Text cleaning & preprocessing

---

📦 Installation

pip install vaderSentiment textblob wordcloud pandas matplotlib seaborn
python -m textblob.download_corpora

---

📁 Dataset

File	Description
IPL_Match_Highlights_Commentary.csv	IPL commentary with Team, Over, Score columns

Columns used:

• Commentary — Raw match commentary text
• Team — Team name
• score — Ball outcome (4, 6, W, dot etc.)

---

🚀 How to Run

Option A — Google Colab (Recommended)

1. Open IPL_SentimentAnalysis.ipynb in Google Colab
2. Upload IPL_Match_Highlights_Commentary.csv to Google Drive
3. Update file path in Step 3
4. Run all cells

Option B — Local

git clone https://github.com/rakesh4407/ipl-sentiment-analysis
cd ipl-sentiment-analysis
pip install -r requirements.txt
jupyter notebook IPL_SentimentAnalysis.ipynb

---

💡 Key Insights

• 🟢 IPL commentary is predominantly Positive — reflecting exciting gameplay
• 🔴 Wicket deliveries generate highest Negative sentiment scores
• 🟡 Dot balls trend toward Neutral sentiment
• 🏏 Boundary (4s & 6s) commentary scores highest Positive compound
• 🎯 VADER outperforms TextBlob for sports commentary analysis

---

🔬 How It Works

Raw Commentary Text
      ↓
Text Cleaning (lowercase, remove special chars)
      ↓
VADER Analysis → Compound Score → Sentiment Label
      ↓
TextBlob Analysis → Polarity + Subjectivity
      ↓
Visualization (Charts, WordClouds, Heatmaps)
      ↓
Insights & Conclusions

---

👨‍💻 Author

Rakesh G

BCA (H) — Artificial Intelligence & Data Science
K.R. Mangalam University, New Delhi | CGPA: 9.22/10
Dean's Award Recipient | IBM Certified Data Scientist

---

🏷️ Topics

python nlp sentiment-analysis vader textblob ipl cricket wordcloud text-analysis data-science machine-learning sports-analytics

---

⭐ If you found this useful, please star this repository!