TikTok Refugees Related Posts Data Dashboard

Overview

This Streamlit-based dashboard provides interactive visualizations for analyzing posts related to "TikTok Refugees." It allows users to explore trends, category distributions, and geographic insights based on uploaded data.

Features

• Data Upload: Users can upload a CSV file for analysis.
• Category Distribution: Visualize the distribution of posts across different categories.
• Time Trends: Track topic trends over time with interactive line charts.
• Geographic Analysis: View post distribution on a province-level heatmap or choropleth map.
• Post Exploration: Filter and examine posts by category and date.

Installation

Ensure you have Python installed, then install the required dependencies:

pip install streamlit pandas numpy folium streamlit-folium plotly requests

Running the App

To launch the dashboard, run:

streamlit run app.py

Online Demo

You can access the online version of this dashboard here:
🔗 TikTok Refugees Data Dashboard
After opening the link in your browser, you can use the examplecsv file from this repository to display data.

Data Management

• The dataset consists of all posts retrieved from Rednote using the keyword "TikTok refugee" between January 2, 2025, and February 19, 2025.

• The categorization of posts was performed using a locally deployed Llama 3.2-8B model.

Data Requirements

The uploaded CSV file should contain the following columns:

• time: Timestamp in milliseconds
• title: Post title
• desc: Post description
• category: Post category (mapped to readable names)
• ip_location: Geographic location (province-level data)

Data Coverage

• Time Range: The dataset covers posts from January 2, 2025, to February 19, 2025.
• Total Posts: The total number of posts available in the dataset will be displayed in the dashboard after uploading the file.

Usage Guide

1. Upload your CSV file via the sidebar.
2. Navigate between tabs:
   • Topic Trend Analysis: View post trends categorized by topic.
   • Geographic Distribution: Explore post distribution across regions.
3. Customize visualizations by selecting specific categories or dates.

Dependencies

• Streamlit
• Pandas
• NumPy
• Folium
• Plotly
• Requests

License

This project is open-source under the MIT License.