Spaces:
Running
Running
| import streamlit as st | |
| from transformers import AutoTokenizer, AutoModelForSequenceClassification | |
| import torch | |
| import re | |
| import pandas as pd | |
| import googleapiclient.discovery | |
| import plotly.express as px | |
| # Load the BERT tokenizer and model | |
| tokenizer = AutoTokenizer.from_pretrained("nlptown/bert-base-multilingual-uncased-sentiment") | |
| model = AutoModelForSequenceClassification.from_pretrained("nlptown/bert-base-multilingual-uncased-sentiment") | |
| # Set up the YouTube API service | |
| api_service_name = "youtube" | |
| api_version = "v3" | |
| DEVELOPER_KEY = "AIzaSyC4Vx8G6nm3Ow9xq7NluTuCCJ1d_5w4YPE" # Replace with your actual API key | |
| youtube = googleapiclient.discovery.build(api_service_name, api_version, developerKey=DEVELOPER_KEY) | |
| # Function to fetch comments for a video ID | |
| def scrape_comments(video_id): | |
| request = youtube.commentThreads().list( | |
| part="snippet", | |
| videoId=video_id, | |
| maxResults=100 | |
| ) | |
| response = request.execute() | |
| comments = [] | |
| for item in response['items']: | |
| comment = item['snippet']['topLevelComment']['snippet'] | |
| comments.append([ | |
| comment['textDisplay'] | |
| ]) | |
| comments_df = pd.DataFrame(comments, columns=['comment']) | |
| # df.head(10). | |
| return comments_df | |
| # Function to fetch video metadata using YouTube API | |
| def fetch_video_info(video_id): | |
| video_id = extract_video_id(video_id) | |
| request = youtube.videos().list( | |
| part="snippet", | |
| id=video_id | |
| ) | |
| response = request.execute() | |
| if response['items']: | |
| video_info = response['items'][0]['snippet'] | |
| channel_name = video_info['channelTitle'] | |
| video_title = video_info['title'] | |
| return channel_name, video_title | |
| else: | |
| raise ValueError("Video not found") | |
| # Function to extract video ID from YouTube URL | |
| def extract_video_id(video_url): | |
| match = re.search(r'(?<=v=)[\w-]+', video_url) | |
| if match: | |
| return match.group(0) | |
| else: | |
| st.error("Invalid YouTube video URL") | |
| # Function to fetch YouTube comments for a video ID | |
| def fetch_comments(video_id): | |
| # Example using youtube-comment-scraper-python library | |
| comments_df = scrape_comments(video_id) | |
| return comments_df | |
| # Function to analyze sentiment for a single comment | |
| def analyze_sentiment(comment): | |
| tokens = tokenizer.encode(comment, return_tensors="pt", max_length=512, truncation=True) | |
| # input_ids = tokens['input_ids'] | |
| # attention_mask = tokens['attention_mask'] | |
| # result = model(input_ids, attention_mask=attention_mask) | |
| result = model(tokens) | |
| sentiment_id = torch.argmax(result.logits) + 1 | |
| if(sentiment_id > 3): | |
| sentiment_label = "Positive" | |
| elif(sentiment_id < 3): | |
| sentiment_label = "Negative" | |
| else: | |
| sentiment_label = "Neutral" | |
| return sentiment_label | |
| def main(): | |
| st.title("YouTube Comments Sentiment Analysis") | |
| # Create sidebar section for app description and links | |
| st.sidebar.title("Comment Feel") | |
| st.sidebar.write("Welcome to the YouTube Comments Sentiment Analysis App 🎥") | |
| st.sidebar.write(""" | |
| **Description** 📝 | |
| This project utilizes a pre-trained sentiment analysis model based on BERT and TensorFlow to analyze the sentiment of comments from a YouTube video. Users can input a YouTube video URL, fetch related comments, and determine their sentiments (positive, negative, or neutral). | |
| Input a valid YouTube video URL in the provided text box 🔗. | |
| Click "Extract Comments and Analyze" to fetch comments and analyze sentiments 🔄. | |
| View sentiment analysis results via pie and bar charts 📊. | |
| Credits 🌟 | |
| Coder: Aniket Panchal | |
| GitHub: https://github.com/Aniket2021448 | |
| Contact 📧 | |
| For any inquiries or feedback, please contact aniketpanchal1257@gmail.com | |
| """) | |
| st.sidebar.write("Feel free to check out my other apps :eyes:") | |
| with st.sidebar.form("app_selection_form"): | |
| st.write("Select an App:") | |
| app_links = { | |
| "Movie-mind": "https://movie-mind.streamlit.app/", | |
| "find-fake-news": "https://find-fake-news.streamlit.app/" | |
| } | |
| selected_app = st.selectbox("Choose an App", list(app_links.keys())) | |
| submitted_button = st.form_submit_button("Go to App") | |
| # Handle form submission | |
| if submitted_button: | |
| selected_app_url = app_links.get(selected_app) | |
| if selected_app_url: | |
| st.sidebar.success("Redirected successfully!") | |
| st.markdown(f'<meta http-equiv="refresh" content="0;URL={selected_app_url}">', unsafe_allow_html=True) | |
| # Dropdown menu for other app links | |
| st.sidebar.write("In case the apps are down, because of less usage") | |
| st.sidebar.write("Kindly reach out to me @ aniketpanchal1257@gmail.com") | |
| st.write("Enter a YouTube video link below: :movie_camera:") | |
| video_url = st.text_input("YouTube Video URL:") | |
| if st.button("Extract Comments and Analyze"): | |
| video_id = extract_video_id(video_url) | |
| if video_id: | |
| comments_df = fetch_comments(video_id) | |
| comments_df['sentiment'] = comments_df['comment'].apply(lambda x: analyze_sentiment(x[:512])) | |
| sentiment_counts = comments_df['sentiment'].value_counts() | |
| channel_name, video_title = fetch_video_info(video_id) | |
| st.write(f"**Channel Name:** {channel_name}") | |
| st.write(f"**Video Description:** {video_title}") | |
| st.write("Based on top :100: comments from this video") | |
| # Create pie chart | |
| st.write("Pie chart representation :chart_with_upwards_trend:") | |
| fig_pie = px.pie(values=sentiment_counts.values, names=sentiment_counts.index, title='Sentiment Distribution') | |
| st.plotly_chart(fig_pie, use_container_width=True) | |
| # Create bar chart | |
| st.write("Bar plot representation :bar_chart:") | |
| fig_bar = px.bar(x=sentiment_counts.index, y=sentiment_counts.values, labels={'x': 'Sentiment', 'y': 'Count'}, title='Sentiment Counts') | |
| st.plotly_chart(fig_bar) | |
| if __name__ == "__main__": | |
| main() | |