Update from Colab

app.py

@@ -1,60 +1,71 @@
-import streamlit as st
+import gradio as gr
 import pandas as pd
-import joblib
 import re
-import string
 from sklearn.feature_extraction.text import TfidfVectorizer
 from sklearn.linear_model import LogisticRegression
+import string
+import numpy as np
+
 
-@st.cache_data
-def load_data():
-    df = pd.read_csv('car_rental_feedback_sentiment_training.csv')
-    return df
+
+# Load data
+df = pd.read_csv('car_rental_feedback_sentiment.csv - Copy (1).csv')  # Make sure filename matches
 
 def preprocess_text(text):
+    if not isinstance(text, str):
+        return ""
     text = text.lower()
     text = text.translate(str.maketrans('', '', string.punctuation))
     text = re.sub('\s+', ' ', text).strip()
     return text
 
-@st.cache_resource
-def train_model(df):
-    df['cleaned_review'] = df['review'].apply(preprocess_text)
-    vectorizer = TfidfVectorizer(max_features=1000)
-    X = vectorizer.fit_transform(df['cleaned_review'])
-    y = df['sentiment_value']
-    model = LogisticRegression()
-    model.fit(X, y)
-    return vectorizer, model
-
-def predict_sentiment(text, vectorizer, model):
-    cleaned_text = preprocess_text(text)
-    text_vector = vectorizer.transform([cleaned_text])
-    prediction = model.predict(text_vector)[0]
-    probabilities = model.predict_proba(text_vector)[0]
-    confidence = probabilities.max()
-    feature_names = vectorizer.get_feature_names_out()
-    sorted_indices = text_vector.toarray().argsort()[0][::-1]
-    top_keywords = [feature_names[i] for i in sorted_indices if text_vector[0,i] > 0][:5]
-    sentiment_label = 'Positive' if prediction == 1 else 'Negative' if prediction == -1 else 'Neutral'
-    return sentiment_label, prediction, round(confidence * 100, 2), top_keywords
-
-st.title("🚗 Car Rental Feedback Sentiment Analyzer")
-st.write("Enter your car rental review below. The app will predict the sentiment (Positive, Neutral, Negative) based on your data.")
-
-df = load_data()
-vectorizer, model = train_model(df)
-
-user_input = st.text_area("Your Car Rental Review", "")
-
-if st.button("Analyze"):
-    if len(user_input.strip()) < 10:
-        st.warning("Please enter a more detailed review.")
-    else:
-        sentiment_label, prediction, confidence, keywords = predict_sentiment(user_input, vectorizer, model)
-        st.markdown("---")
-        st.subheader("Result")
-        st.write(f"**Sentiment:** {sentiment_label} ({prediction})")
-        st.write(f"**Confidence:** {confidence}%")
-        st.write(f"**Key Features:** {', '.join(keywords) if keywords else 'None detected'}")
-        st.markdown("---")
+# Check if required columns exist
+if not all(col in df.columns for col in ['review', 'sentiment_value']):
+    raise ValueError("CSV file must contain 'review' and 'sentiment_value' columns")
+
+df['cleaned_review'] = df['review'].apply(preprocess_text)
+vectorizer = TfidfVectorizer(max_features=1000)
+X = vectorizer.fit_transform(df['cleaned_review'])
+y = df['sentiment_value']
+model = LogisticRegression(max_iter=1000)
+model.fit(X, y)
+
+def predict_sentiment(text):
+    try:
+        cleaned_text = preprocess_text(text)
+        text_vector = vectorizer.transform([cleaned_text])
+        prediction = model.predict(text_vector)[0]
+        probabilities = model.predict_proba(text_vector)[0]
+        confidence = probabilities.max()
+        sentiment_label = 'Positive' if prediction == 1 else 'Negative' if prediction == -1 else 'Neutral'
+        return f"{sentiment_label} (confidence: {confidence*100:.1f}%)"
+    except Exception as e:
+        return f"Error processing your input: {str(e)}"
+
+description = """
+<div style="display: flex; align-items: center; gap: 24px; background: #fff; padding: 16px 32px; border-radius: 14px; box-shadow: 0 2px 12px rgba(0,0,0,0.08); width: fit-content;">
+  <img src="https://upload.wikimedia.org/wikipedia/en/thumb/c/c5/Vellore_Institute_of_Technology_seal_2017.svg/300px-Vellore_Institute_of_Technology_seal_2017.svg.png" alt="VIT Logo" style="height: 48px; width: auto; border-radius: 8px; background: #f5f7fa; padding: 6px; box-shadow: 0 1px 6px rgba(0,0,0,0.05); transition: transform 0.2s;" onmouseover="this.style.transform='scale(1.08)'" onmouseout="this.style.transform='scale(1)'"/>
+  <img src="https://www.ibm.com/brand/experience-guides/developer/8f4e3cc2b5d52354a6d43c8edba1e3c9/02_8-bar-reverse.svg" alt="IBM Logo" style="height: 48px; width: auto; border-radius: 8px; background: #f5f7fa; padding: 6px; box-shadow: 0 1px 6px rgba(0,0,0,0.05); transition: transform 0.2s;" onmouseover="this.style.transform='scale(1.08)'" onmouseout="this.style.transform='scale(1)'"/>
+</div>
+<br>
+<b style="font-size: 1.3rem; letter-spacing: 1px; color: #222; font-family: 'Segoe UI', Arial, sans-serif;">
+  Car Rental Feedback Analyzer
+</b>
+<p>Enter your car rental review below. The app will predict the sentiment (Positive, Neutral, Negative).</p>
+"""
+
+iface = gr.Interface(
+    fn=predict_sentiment,
+    inputs=gr.Textbox(label="Enter car rental review", placeholder="Type your car rental experience here..."),
+    outputs=gr.Textbox(label="Sentiment Prediction"),
+    title="🚗 Car Rental Feedback Analyzer",
+    description=description,
+    examples=[
+        ["The car was clean and the staff was friendly"],
+        ["Terrible experience with late delivery"],
+        ["Average service, nothing special"]
+    ]
+)
+
+# Test locally in Colab first
+iface.launch()