SMS Spam Detection: Combined Model Card

Models

1. Multinomial Naive Bayes

• Type: MultinomialNB
• Library: scikit-learn
• Description: A Naive Bayes classifier for multinomially distributed data, commonly used for text classification tasks.
• Training Data: SMS Spam Collection dataset (train.csv), preprocessed and vectorized using CountVectorizer.
• Features: Bag-of-words (unigrams), stopwords removed.
• Target: label (0: ham, 1: spam)
• Accuracy: {{ accuracy_score(tahmin, y_test) }}
• Date Trained: {{ datetime.now().strftime("%Y-%m-%d") }}

2. Decision Tree Classifier

• Type: DecisionTreeClassifier
• Library: scikit-learn
• Description: A decision tree classifier for binary classification of SMS messages.
• Training Data: SMS Spam Collection dataset (train.csv), preprocessed and vectorized using CountVectorizer.
• Features: Bag-of-words (unigrams), stopwords removed.
• Target: label (0: ham, 1: spam)
• Accuracy: {{ accuracy_score(tahmin3, y_test) }}
• Date Trained: {{ datetime.now().strftime("%Y-%m-%d") }}

Preprocessing

• Lowercasing all text
• Removing punctuation, digits, and newlines
• Stopwords removed during vectorization

Evaluation Metric

• Accuracy on test set

Notes

• Models saved using joblib.
• For further evaluation, consider precision, recall, and F1-score.