This project implements an Automated Essay Scoring (AES) system using machine learning techniques, specifically LSTM layers. The goal is to develop a model that can evaluate essays based on predefined criteria and produce a score that approximates human scoring. This project serves as a foundation for assessing textual content, a useful application in educational technology and online learning platforms.
Automated Essay Scoring (AES) is a fast-growing area in natural language processing and machine learning, aiming to assist educators by providing a scalable, objective, and rapid assessment solution for written content. This project applies a Long Short-Term Memory (LSTM) model, an effective model for handling sequential data like text, to predict essay scores.
- Text Preprocessing: Includes cleaning, tokenization, and vectorization of essays for effective modeling.
- LSTM Model Architecture: Designed to handle sequential data with recurrent layers to capture the contextual meaning.
- Evaluation Metrics: The model performance is evaluated based on mean squared error (MSE) to quantify the difference between predicted and actual scores.
main.ipynb: Jupyter notebook containing code for data processing, model building, training, and evaluation.data/: Folder to store datasets used for training and testing the AES model.models/: Directory for saving trained model files.README.md: Project description and setup instructions.
-
Clone the Repository
git clone https://github.com/yourusername/automated-essay-scoring.git cd automated-essay-scoring