NEGATIVE SOCIAL MEDIA COMMENTS CLASSIFICATION

Our project focuses on building an end-to-end system to automatically crawl Facebook comments data and classify them into four labels: Hate speech, Personal Attack, Discrimination, Others.

Main Fuction

• Scrape comments with related information such as:
  • Comment - text
  • User
  • Nametag (like @name_tag)
  • Check if it is spam (based on user-defined demand)
  • Via https://github.com/boo283/Facebook_comment_crawler.git
• Preprocess data:
  • Follow these steps:
• Make classification and visualize the results:
  • Use HSD model (base on ML model) to predict
  • Visualize data using Streamlit

Installation

• Clone the repository git clone https://github.com/boo283/Facebook_comment_crawler.git git clone https://github.com/boo283/NEGATIVE-SOCIAL-MEDIA-COMMENTS-CLASSIFICATION.git
• Install dependencies: pip install -r requirements.txt

Usage

1. Clone this repository
2. Open this repository and add some information:

• Crawler
• In folder "configuration":
  • config.py:
    • In function configure_driver(), replace by the path to your chrome driver, which could be download at ref: https://googlechromelabs.github.io/chrome-for-testing/#stable
• In main folder:
  • crawl.py:
    • Just type your Facebook account in the Login info part in main function.
    • Choose your destination to save crawled data
• App.py:
  • Create kafka-topic "StreamComments"
  • Adjust paths in python files

3. Cd to folder and run script: python app.py
4. Simply enter the Facebook URL post, Account information and choose Featch and analyze The result will like:

Contact:

• LinkedIn: https://www.linkedin.com/in/phutrungnguyen283/
• Facebook: https://www.facebook.com/ngphtrungboo
• Email: trung280302@gmail.com

Feel free to adjust my code, practice makes perfect ❤️❤️❤️