import streamlit as st

st.title("📊 Data Collection & Preprocessing")

st.markdown("""

### 🔍 **YouTube Comment Extraction**

✅ Used **YouTube Data API v3** via the **requests** library and **API Key** authentication

📹 Extracted comments from 150+ video IDs across balanced sentiment categories:

🟢 **Positive:** "heartwarming reunion", "feel good moments"

🔴 **Negative:** "controversial incident", "bad customer experience"

⚪ **Neutral:** "tech reviews", "travel vlogs"


📈 Gathered over **15,000** top-level comments using pagination (nextPageToken)

💾 Data saved into a **JSON** format with keys: **video_id** and **comment**



---

### 🧹 **Text Preprocessing Steps**

✅ Removed HTML tags, emojis, special characters, and links

🔡 Converted text to lowercase

🧽 Removed stopwords and non-alphabetic tokens

🧠 Applied **tokenization**


The cleaned text was used to train our **sentiment classifier** in the next step.


---

""")

if st.button("Next"):
    st.switch_page("pages/2_Model_Training.py")

if st.button("Go to Home"):
    st.switch_page("app.py")