Update app.py

app.py

@@ -1,84 +1,110 @@
-# =====================================================
-# 🌐 Multi-lingual Sentiment Analyzer (English + Persian)
-# =====================================================
+# ===============================================================
+# 🌍 Sentiment Analysis App — English & Persian (Gradio Deployment)
+# ===============================================================
 
 import gradio as gr
 import joblib
-import numpy as np
 import shap
-import os
-
-# -----------------------------------------------------
-# ✅ Load Models and Vectorizers
-# -----------------------------------------------------
-english_model = joblib.load("best_model.pkl")
-english_vectorizer = joblib.load("tfidf_vectorizer.pkl")
+import numpy as np
 
-persian_model = joblib.load("logistic_regression.pkl")
-persian_vectorizer = joblib.load("tfidf_vectorizer_persian.pkl")
+# ------------------------------------------------------------
+# 1️⃣ Load pre-trained models and vectorizers
+# ------------------------------------------------------------
+eng_model = joblib.load("best_model.pkl")
+eng_vectorizer = joblib.load("tfidf_vectorizer.pkl")
 
+per_model = joblib.load("logistic_regression.pkl")
+per_vectorizer = joblib.load("tfidf_vectorizer_persian.pkl")
 
-# -----------------------------------------------------
-# ✅ Label mapping
-# -----------------------------------------------------
-label_map = {0: "Negative", 1: "Neutral", 2: "Positive"}
+# ------------------------------------------------------------
+# 2️⃣ Define class labels
+# ------------------------------------------------------------
+class_names = ["Negative", "Neutral", "Positive"]
 
-# -----------------------------------------------------
-# ✅ Prediction Function
-# -----------------------------------------------------
-def predict_sentiment(text, lang):
+# ------------------------------------------------------------
+# 3️⃣ Prediction Function
+# ------------------------------------------------------------
+def predict_sentiment(text, language):
     if not text.strip():
-        return "⚠️ Please enter some text.", "", "", ""
-    
-    # Select appropriate model and vectorizer
-    if lang == "English":
-        vectorizer = eng_vectorizer
+        return "⚠️ Please enter some text!", None, None
+
+    if language == "English":
         model = eng_model
+        vectorizer = eng_vectorizer
     else:
-        vectorizer = per_vectorizer
         model = per_model
+        vectorizer = per_vectorizer
+
+    # Vectorize input
+    vec = vectorizer.transform([text])
+    probs = model.predict_proba(vec)[0]
+    pred_class = np.argmax(probs)
+    label = class_names[pred_class]
+    confidence = probs[pred_class]
+
+    # Interpret top words with SHAP
+    explainer = shap.Explainer(model, vectorizer.transform(["sample"]))
+    feature_names = vectorizer.get_feature_names_out()
+    shap_values = explainer(vec)
+
+    # top contributing words
+    shap_vals = shap_values[0].values[:, pred_class]
+    top_indices = np.argsort(-np.abs(shap_vals))[:10]
+    top_words = [feature_names[i] for i in top_indices]
+    top_contribs = shap_vals[top_indices]
+
+    interpretation = {
+        "words": top_words,
+        "contributions": top_contribs.tolist()
+    }
+
+    return f"🎯 **{label}** (confidence: {confidence:.2f})", probs.tolist(), interpretation
 
-    # Vectorize text
-    X = vectorizer.transform([text])
-    pred = model.predict(X)[0]
-    probs = model.predict_proba(X)[0]
-    conf = np.max(probs)
-    sentiment = label_map.get(pred, "Unknown")
-
-    # SHAP explanation
-    try:
-        explainer = shap.LinearExplainer(model, X, feature_dependence="independent")
-        shap_values = explainer.shap_values(X)
-        shap_html = shap.plots.text(explainer, X, display=False)
-    except Exception:
-        shap_html = "<p>⚠️ No explanation available for this input.</p>"
-
-    return f"Prediction: {sentiment}", f"Confidence: {conf:.3f}", shap_html, ""
-
-# -----------------------------------------------------
-# ✅ Build Gradio Interface
-# -----------------------------------------------------
-with gr.Blocks(theme=gr.themes.Soft()) as app:
-    gr.Markdown("<h2 style='text-align:center;'>🌍 Multi-lingual Sentiment (English + Persian)</h2>")
-    
-    with gr.Row():
-        comment = gr.Textbox(label="Comment", placeholder="Type your comment here...")
-        lang = gr.Radio(["English", "Persian"], label="Language", value="English")
-    
-    predict_btn = gr.Button("Predict", variant="primary")
-    
-    output1 = gr.Textbox(label="Prediction")
-    output2 = gr.Textbox(label="Confidence")
-    output3 = gr.HTML(label="Explanation")
-
-    predict_btn.click(
-        predict_sentiment,
-        inputs=[comment, lang],
-        outputs=[output1, output2, output3, gr.Textbox(visible=False)]
-    )
-
-# -----------------------------------------------------
-# ✅ Launch App
-# -----------------------------------------------------
+
+# ------------------------------------------------------------
+# 4️⃣ Gradio Interface
+# ------------------------------------------------------------
+def gradio_ui(text, language):
+    pred, probs, interp = predict_sentiment(text, language)
+
+    if not probs:
+        return pred, None, None
+
+    # Confidence Bar Plot
+    bar_plot = {cls: float(p) for cls, p in zip(class_names, probs)}
+
+    # Word Contribution Table
+    word_table = None
+    if interp:
+        word_table = {
+            "Word": interp["words"],
+            "SHAP Impact": interp["contributions"]
+        }
+
+    return pred, bar_plot, word_table
+
+
+# ------------------------------------------------------------
+# 5️⃣ Gradio Layout
+# ------------------------------------------------------------
+with gr.Blocks(theme=gr.themes.Soft()) as interface:
+    gr.Markdown("## 🌍 Sentiment Analysis (English & Persian)  
+    Select language, enter text, and view predictions with interpretable SHAP insights!")
+
+    lang = gr.Radio(["English", "Persian"], label="Select Dataset", value="English")
+    text = gr.Textbox(label="Enter your text here", placeholder="Type an English or Persian comment...")
+
+    output_pred = gr.Markdown(label="Prediction")
+    output_bar = gr.BarPlot(label="Confidence per Class")
+    output_table = gr.Dataframe(label="Top Influential Words", headers=["Word", "SHAP Impact"])
+
+    btn = gr.Button("🔍 Analyze Sentiment")
+
+    btn.click(fn=gradio_ui, inputs=[text, lang], outputs=[output_pred, output_bar, output_table])
+
+# ------------------------------------------------------------
+# 6️⃣ Launch App
+# ------------------------------------------------------------
 if __name__ == "__main__":
-    app.launch()
+    interface.launch(share=True)
+