Text Analytics

( Sentiment Classification )

💭

ABSTRACT :

---

This is a text classification task - sentiment classification.
Every document (a line in the data file) is a sentence extracted from social media (blogs).
The goal is to classify the sentiment of each sentence into "positive" or "negative".
Accuracy of this model: 97.91%

DATASET:

---

• The training data contains 7086 sentences, already labeled with 1 (positive sentiment) or 0 (negative sentiment).
• The test data contains 33052 sentences that are unlabeled.
• The submission should be a .txt file with 33052 lines.
• In each line, there should be exactly one integer, 0 or 1, according to the classification results.

SOLUTION APPROACH:

---

 Text Analytics
     NLP + ML (Naive Bayes Theoerm)  
     Ensemble Learning

STEPS:

---

1. EDA and Pre-processing

• Bag-of-Words method
• Feature Extraction
• Removing the low frequency words
• Count the features
• Remove Stop Words
• Stemming

2. Balancing the unbalanced data classes

3. Naive Bayes Model for sentiment Classification

• Dataset segmentation
• Model building
• Prediction test
• Classification report

DATA VISUALIZATION:

---

• Sample Data:

• Positive and Negative Class-wise Data Distribution Bar Plot :

• The Occurances of Features (Column-wise) :

• Positive and Negative Class-wise Data Distribution After Balancing Data-Classes:

RESULTS:

---

• The Table with Results:

• Confusion Matrix

• The Prediction Accuracy Score for Training Data : 99.37%
• The Prediction Accuracy Score for Testing Data : 97.91%