This project explores multimodal sentiment analysis for architecture-related content on social media platforms. By leveraging text, image, and video data, we aim to assess public sentiment and engagement towards architectural designs, styles, and trends.

• Multimodal Data Processing: Combines text (comments, captions), images, and video features for sentiment prediction.
• Deep Learning-Based Sentiment Analysis: Utilizes LLMs for text and vision models for image/video understanding.
• Engagement Score Prediction: Analyzes user interactions to determine sentiment-driven engagement.
• Dataset Integration: Supports Instagram metadata for sentiment analysis.
• Visualization Tools: Provides sentiment distribution insights and engagement trends.

The project works with datasets from Instagram, containing:

• Text: Video titles, descriptions, hashtags, and user comments.
• Visual: Extracted frames from videos.
• Engagement Metrics: Likes, views, shares, and comments.

Clone the repository:

git clone https://github.com/yourusername/Multimodal-Sentiment-Analysis-for-Architecture-on-Social-Media.git
cd Multimodal-Sentiment-Analysis-for-Architecture-on-Social-Media