random_forest.py

from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.ensemble import RandomForestClassifier
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score, classification_report
import pandas as pd

# Load dataset
df = pd.read_csv("dataset.csv")

# Split dataset into training and validation sets
train_texts, val_texts, train_labels, val_labels = train_test_split(
    df["column_name"].tolist(),
    df["is_hipaa_sensitive"].tolist(),
    test_size=0.2,
    random_state=42,
    stratify=df["is_hipaa_sensitive"]
)

# Convert text into TF-IDF features
vectorizer = TfidfVectorizer(
    ngram_range=(1, 2),     # unigrams + bigrams
    max_features=20000     # helps Random Forest performance
)

X_train = vectorizer.fit_transform(train_texts)
X_val = vectorizer.transform(val_texts)

# Train Random Forest model
rf = RandomForestClassifier(
    n_estimators=500,
    max_depth=None,
    min_samples_split=2,
    min_samples_leaf=1,
    random_state=42,
    n_jobs=-1,
    class_weight="balanced"   # important for HIPAA imbalance
)

rf.fit(X_train, train_labels)

# Make predictions
pred_labels = rf.predict(X_val)

# Calculate accuracy
accuracy = accuracy_score(val_labels, pred_labels)
print(f"Random Forest Accuracy: {accuracy * 100:.4f}%")

# Print classification report
print("\nClassification Report:\n",
      classification_report(val_labels, pred_labels, target_names=["Non-Sensitive", "Sensitive"]))

# Function to test new examples
def predict_tf_idf(texts):
    X_test = vectorizer.transform(texts)
    predictions = rf.predict(X_test)
    return ["Sensitive" if p == 1 else "Non-sensitive" for p in predictions]

# Test on new column names
test_texts = [
    "birthDate", "birth_year", "country",
    "DATE_BIRTH", "color", "food", "jwtToken"
]

predictions = predict_tf_idf(test_texts)

# Print predictions
print("\nPredictions:")
for text, pred in zip(test_texts, predictions):
    print(f"{text}: {pred}")