Skip to content

ishitaagl20/Authentication-System-NLP

Repository files navigation

AI Powered Authentication System

This project implements a Conversational AI Authentication System with two distinct modules:

  1. Password Strength Evaluation: Evaluates the strength of user passwords based on a trained model.
  2. Text-Based Conversational Authentication: Authenticates users based on dynamically generated questions from a descriptive input.

Features

1. Password Strength Evaluation (Module 1)

  • Uses a RandomForestClassifier and TF-IDF Vectorizer to predict the strength of a password.
  • The system evaluates password strength based on factors like length, uppercase/lowercase letters, numbers, special characters, etc.

2. Text-Based Conversational Authentication (Module 2)

  • Takes a descriptive input from the user (e.g., details about the user’s background), instead of security questions.
  • Generates questions dynamically from the input using an AI model (Gemini-Pro).
  • Validates the answers provided by the user for authentication.
  • This is extensible to scrapping data from user activity and then posing questions.

Installation

Clone the repository:

git clone https://github.com/your-repo/conversational-ai-authentication.git