Smart Teaching Assistant

Abstract

The Smart Teaching Assistant (STA) is an AI-powered educational tool designed to automate the creation and assessment of academic questions. It uses Natural Language Processing (NLP) and machine learning techniques, especially Recurrent Neural Networks (RNNs), to generate relevant, diverse questions from provided text and evaluate student answers effectively. The system enables teachers to upload learning material, from which questions such as multiple-choice, short answer, and descriptive types are generated. It also evaluates answers based on semantic similarity with predefined responses, providing accurate feedback and scores. STA addresses major challenges in traditional education systems, such as time constraints, inconsistencies in evaluation, and lack of personalization. With a modular design supporting continuous model improvement via user feedback, the system ensures scalability, adaptability.