Sentiment Analysis

What is Sentiment Analysis?

Sentiment analysis (or opinion mining) is an NLP technique that determines the emotional tone behind a piece of text — whether it’s positive, negative, or neutral. It’s how brands gauge public opinion at scale, analyzing thousands of reviews, social media posts, and customer feedback instantly.

Sentiment analysis extracts emotional tone from text. Know what people really feel about products, brands, and services. “This product is amazing” vs. “This product is terrible” — sentiment models classify these automatically, turning unstructured opinions into quantifiable data.

Sentiment Analysis Approaches

Method	Use Case	Accuracy
Lexicon-based	Quick baseline	Medium
ML-based	General purpose	High
LLM-based	Complex nuance	Highest

Implementation Examples

1. VADER (Social Media, Reviews)

from vaderSentiment.vaderSentiment import SentimentIntensityAnalyzer

analyzer = SentimentIntensityAnalyzer()

texts = [
    "This product is absolutely amazing! Best purchase ever.",
    "Terrible service. Never buying again.",
    "It's okay, not great but not bad either."
]

for text in texts:
    scores = analyzer.polarity_scores(text)
    print(f"{text[:40]:40} → {scores['compound']:.2f}")

# Output:
# This product is absolutely amazing! Best p... →  0.93 (Positive)
# Terrible service. Never buying again.       → -0.82 (Negative)
# It's okay, not great but not bad either.    →  0.00 (Neutral)

2. Transformers (Fine-Grained)

from transformers import pipeline

sentiment_analyzer = pipeline("sentiment-analysis")

text = "The interface is user-friendly but the pricing is too high."

result = sentiment_analyzer(text)
# [{'label': 'NEGATIVE', 'score': 0.75}]

3. Aspect-Based (Understand Specific Features)

# Analyze sentiment per aspect
def analyze_product_review(review):
    aspects = {
        'battery': extract_sentence(review, 'battery'),
        'camera': extract_sentence(review, 'camera'),
        'screen': extract_sentence(review, 'screen')
    }
    
    return {
        'overall': analyze(review),
        'battery': analyze(aspects['battery']),
        'camera': analyze(aspects['camera']),
        'screen': analyze(aspects['screen'])
    }

# Output:
# {'overall': 'Positive', 'battery': 'Positive', 'camera': 'Negative', 'screen': 'Neutral'}

Scraping + Sentiment Pipeline

# Scrape reviews → Analyze sentiment → Dashboard
reviews = scrape_product_reviews('https://example.com/product')

sentiment_data = []
for review in reviews:
    score = analyzer.polarity_scores(review['text'])
    sentiment_data.append({
        'product': review['product'],
        'rating': review['rating'],
        'sentiment_score': score['compound'],
        'sentiment_label': classify(score['compound'])
    })

# Aggregate insights
df = pd.DataFrame(sentiment_data)
print(f"Average Sentiment: {df['sentiment_score'].mean():.2f}")
print(f"Positive Reviews: {(df['sentiment_label'] == 'Positive').sum()}")

Business application: Track brand sentiment over time, compare competitor sentiment, identify pain points from negative reviews, and validate marketing campaign effectiveness.

Sentiment Analysis

What is Sentiment Analysis?