# ---------- Sentiment Analyzer Pro (With CSV Export and Language Support) ---------- import streamlit as st import pandas as pd import time from transformers import pipeline from io import StringIO import requests # -------------------- Configurations -------------------- st.set_page_config( page_title="Sentiment Analyzer Pro", page_icon="🧠", layout="wide", ) # -------------------- Helper Functions -------------------- @st.cache_resource def load_pipeline(language='en'): if language == 'ur': return pipeline("sentiment-analysis", model="urduhack/bert-base-urdu-cased-sentiment") return pipeline("sentiment-analysis") classifier = load_pipeline() def analyze_sentiment(text, language): result = classifier(text)[0] return result["label"], result["score"] def create_csv(data): df = pd.DataFrame(data) csv = df.to_csv(index=False) return csv def save_csv(data, filename="sentiment_analysis.csv"): csv = create_csv(data) st.download_button( label="Download CSV", data=csv, file_name=filename, mime="text/csv", ) def submit_contact_form(name, email, message): # [Optional] You can connect this with Google Forms or Email APIs st.success(f"✅ Thank you {name}, your message has been received!") def show_footer(): st.markdown("---") st.markdown( "

Š 2025 Sentiment Analyzer Pro | Built with ❤ī¸ using Hugging Face and Streamlit

", unsafe_allow_html=True, ) # -------------------- Sidebar -------------------- st.sidebar.title("📚 Navigation") page = st.sidebar.radio("Go to", ["Home", "About", "Contact"]) st.sidebar.title("📚 Use Cases (on Home)") use_case = st.sidebar.radio( "Choose a Use Case", ( "General Text", "Social Media Post", "Customer Review", "Email/Message", "Product Description", ) ) st.sidebar.title("🌐 Language") language = st.sidebar.selectbox( "Select Language", ("English", "Urdu"), ) st.sidebar.markdown("---") st.sidebar.info( "This app uses a **Transformers-based model** to predict the **sentiment** " "(Positive, Negative, Neutral) with a **confidence score**." ) st.sidebar.write("Made with ❤ī¸ by M-Abdullah") # -------------------- Page Logic -------------------- if page == "Home": # ------------- HOME PAGE ------------- st.markdown("

🧠 Sentiment Analyzer Pro

", unsafe_allow_html=True) st.markdown("

Analyze emotions in text, reviews, posts, and more!

", unsafe_allow_html=True) st.write("---") st.write(f"### ✍ī¸ Enter your {use_case} below:") user_input = st.text_area("", placeholder=f"Write your {use_case.lower()} here...", height=200) if st.button("🔍 Analyze Sentiment"): if user_input.strip() == "": st.warning("⚠ī¸ Please enter some text to analyze!") else: with st.spinner('Analyzing...'): label, score = analyze_sentiment(user_input, language) time.sleep(1) # Smooth spinner effect # Emoji Reaction emoji = { "POSITIVE": "😊", "NEGATIVE": "😞", "NEUTRAL": "😐" }.get(label.upper(), "🤔") # Result Display st.success(f"**Sentiment:** {label} {emoji}") st.info(f"**Confidence Score:** {score:.2f}") # Chart Section st.write("### 📊 Sentiment Score") st.bar_chart({"Confidence Score": [score]}) # Detailed Explanation st.write("### 📖 Interpretation") if label == "POSITIVE": st.write("✅ This text expresses positive emotions, happiness, satisfaction, or support.") elif label == "NEGATIVE": st.write("⚠ī¸ This text expresses negative emotions, dissatisfaction, criticism, or sadness.") else: st.write("ℹī¸ This text is neutral without strong emotions.") # Download CSV Button save_csv([{"Text": user_input, "Sentiment": label, "Confidence Score": score}], "sentiment_analysis.csv") elif page == "About": # ------------- ABOUT PAGE ------------- st.markdown("

📘 About Sentiment Analyzer Pro

", unsafe_allow_html=True) st.write("---") st.markdown(""" ### ✨ Overview **Sentiment Analyzer Pro** is an advanced natural language processing (NLP) app built using: - 🛠 **Streamlit** for the web app interface - 🧠 **Hugging Face Transformers** for deep learning models This application can detect and classify **sentiments** (Positive, Negative, Neutral) in: - General texts - Social media posts (e.g., Twitter, Facebook) - Customer feedback - Emails and messaging - Product descriptions ### đŸŽ¯ Objectives - Simplify sentiment analysis for all users - Provide fast, reliable, real-time emotional analysis - Useful for businesses, researchers, students, and individuals ### 🚀 Future Enhancements - Multilingual Sentiment Analysis (Urdu, Arabic, French) - Custom Model Uploads - Sentiment Trends over Time """) elif page == "Contact": FORM_URL = "https://docs.google.com/forms/u/0/d/1A95xalIMN6jDN8DCAq3agQGGDaDXP7x1hed-DjHqxow/prefill" FIELD_IDS = { "name": "entry.796411702", # replace with actual ID "email": "entry.796411702", # replace with actual ID "message": "entry.939908203", # Correct ID } # Page Title st.markdown("

đŸ“Ŧ Contact Us

", unsafe_allow_html=True) st.markdown("---") st.markdown(""" ### đŸ’Ŧ We'd Love to Hear From You! Fill out the form below and your message will be saved to our system. """) with st.form(key='contact_form'): name = st.text_input("Your Name") email = st.text_input("Your Email") message = st.text_area("Your Message") submit_button = st.form_submit_button(label='Send Message') if submit_button: if name.strip() == "" or email.strip() == "" or message.strip() == "": st.warning("⚠ī¸ Please fill out all fields before submitting.") else: # Debugging: Check form values st.write(f"Name: {name}, Email: {email}, Message: {message}") # Submit data to Google Form data = { FIELD_IDS["name"]: name, FIELD_IDS["email"]: email, FIELD_IDS["message"]: message, } response = requests.post(FORM_URL, data=data) if response.status_code == 200: st.success(f"✅ Thank you {name}! Your message has been sent successfully.") else: st.error(f"❌ There was a problem sending your message. Error: {response.status_code}") # Footer st.markdown("---") st.markdown("

Made with ❤ī¸ by M-Abdullah | 2025

", unsafe_allow_html=True) # -------------------- Footer -------------------- show_footer()