Building a Simple Sentiment Analyzer with Python (TextBlob & VADER)

Sentiment analysis is the process of determining the emotional tone behind a piece of text. It's used everywhere — social media monitoring, product review analysis, customer feedback systems, and chatbots.

In this tutorial we'll build a command-line sentiment analyzer that classifies input as positive, negative, or neutral using two different NLP libraries and comparing their results.

What We're Building

A Python CLI app that:

Accepts user text input in real time
Runs it through TextBlob and VADER simultaneously
Returns an emoji-tagged classification from both models

Libraries

TextBlob

Provides a simple API returning a polarity score from -1 (very negative) to +1 (very positive).

VADER (Valence Aware Dictionary and sEntiment Reasoner)

Specialised for informal text — social media posts, slang, emojis. Returns a compound score for overall classification.

Setup

bash

pip install textblob vaderSentiment

Download TextBlob corpora:

bash

python -m textblob.download_corpora

Implementation

python

from textblob import TextBlob
from vaderSentiment.vaderSentiment import SentimentIntensityAnalyzer

# Initialise VADER
analyzer = SentimentIntensityAnalyzer()


def analyze_sentiment_textblob(text: str) -> str:
    polarity = TextBlob(text).sentiment.polarity
    if polarity > 0:
        return "Positive 😊"
    elif polarity < 0:
        return "Negative 😡"
    else:
        return "Neutral 😐"


def analyze_sentiment_vader(text: str) -> str:
    compound = analyzer.polarity_scores(text)["compound"]
    if compound >= 0.05:
        return "Positive 😊"
    elif compound <= -0.05:
        return "Negative 😡"
    else:
        return "Neutral 😐"


def main():
    print("\n🧠 Sentiment Analyzer (TextBlob + VADER)")
    print("Type '0' to quit\n")

    while True:
        text = input("How are you feeling? → ")
        if text.strip().lower() == "0":
            print("Exiting. Goodbye! 👋")
            break

        print(f"  TextBlob : {analyze_sentiment_textblob(text)}")
        print(f"  VADER    : {analyze_sentiment_vader(text)}\n")


if __name__ == "__main__":
    main()

Sample Output

text

🧠 Sentiment Analyzer (TextBlob + VADER)
Type '0' to quit

How are you feeling? → I love coding with Python
  TextBlob : Positive 😊
  VADER    : Positive 😊

How are you feeling? → This bug is driving me crazy
  TextBlob : Negative 😡
  VADER    : Negative 😡

How are you feeling? → The weather is okay
  TextBlob : Neutral 😐
  VADER    : Positive 😊

Note how VADER and TextBlob can disagree on borderline cases — VADER is generally better for casual language.

When to Use Which

Scenario	Recommended
Formal text, articles	TextBlob
Tweets, comments, slang	VADER
General purpose	Both + compare

What's Next

This is your entry point into NLP. Once you're comfortable here, explore:

Hugging Face Transformers — state-of-the-art sentiment models
Fine-tuning BERT on custom datasets
Streamlit to turn this into a web app
Batch processing CSV files of reviews

Sentiment analysis is one of the most practical NLP tasks — you'll find it everywhere in real products.