Skip to content

cosmic-invaders/Truth-n-Ruth

Repository files navigation

Truth-n-Ruth

Description

Truth-n-Ruth is an intelligent and powerful tool designed to combat disinformation, hateful content, and offensive material across various types of media. Our app uses advanced natural language processing (NLP) and computer vision techniques to detect fake articles, offensive or hate speech, and hateful memes.

Features

  • Fake Article Detection: Analyze the authenticity of news articles and provide insights into their credibility.

  • Offensive and Hate Speech Detection: Identify and flag offensive language and hate speech within text content.

  • Hateful Meme Recognition: Utilize image analysis to recognize and prevent the sharing of hateful memes.

  • User-Friendly Interface: A sleek and intuitive user interface for easy interaction and seamless experience.

How It Works

Truth-n-Ruth combines the power of machine learning, NLP algorithms, and computer vision models to provide a comprehensive solution for content verification. It follows these steps:

  1. Fake Article Detection:

    • Analyzes the linguistic patterns and context of the article.
    • Evaluates the credibility of sources and cross-references against reliable databases.
    • Flags articles with potential misinformation or fake claims.
  2. Offensive and Hate Speech Detection:

    • Utilizes a comprehensive list of offensive and hate speech keywords and phrases.
    • Employs advanced NLP techniques to assess the sentiment and tone of the text.
    • Provides warnings and suggestions for more respectful communication.
  3. Hateful Meme Recognition:

    • Processes meme images using computer vision models.
    • Identifies visual elements associated with hateful content.
    • Alerts users about harmful or inappropriate memes.

Installation and Usage

  1. Clone this repository:

    git clone https://github.com/yourusername/Truth-n-Ruth.git
    cd Truth-n-Ruth
    streamlit run app.py
    debug yourselves

Releases

No releases published

Packages

No packages published

Languages