International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 05 | May 2024
p-ISSN: 2395-0072
www.irjet.net
Enriching Question Bank using Recurrent Neural Network Harsha Anand PP, Lekha Treesa Titus, Alfeena Sulaiman, Ajumol PA 1Mar Athanasius College of Engineering, Kothamangalam, Kerala, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - —The notable effectiveness of question papers
also offer the flexibility to customize parameters such as question difficulty and topic coverage based on specific educational objectives. This personalized approach enhances the overall quality of assessments and contributes to a more engaging and effective learning experience for students and educators alike. Automated question paper generation using Recurrent Neural Networks (RNNs) faces the challenge of ensuring diversity, relevance, and appropriate difficulty levels in the generated questions, which are essential for effective assessments. This problem is addressed by sophisticated algorithms that mimic human questioncrafting abilities while leveraging the capabilities of RNNs to analyze patterns, detect correlations, and adapt question difficulty based on learning objectives and student capabilities. The preference for RNN-based systems lies in their ability to automate manual question creation, maintain fairness and relevance in assessments, and provide scalable solutions that align with diverse educational contexts and standards
crafted by human experts makes the development of Recurrent Neural Network (RNN) based automated systems a captivating area of study, despite significant advancements in technology. This project aims to create an enriching question paper-generating system, based on RNN, with added features for sorting and personalized recommendations. Automatic question generation for textual inputs is valuable in academics where answering questions helps students to learn and improve their understanding of their field of study. The proposed system automates question collection and provides personalized question papers for students based on performance which helps them to excel in academics. Providing practice question paper sets inculcates various domains of the subject familiarizing with core principles resulting in increasing curiosity and a better understanding of the subject. This comprehensive approach ensures a fair and effective exam preparation experience, fostering continuous improvement while respecting ethical considerations.
2. RELATED WORKS
Key Words: Automated question paper generation system, Recurrent Neural Networks (RNN), Question Bank, LSTM, DuerQues, PyPDF
2.1 Recurrent Neural Network Recurrent Neural Networks (RNNs) stand out in the realm of Artificial Intelligence and Machine Learning for their ability to process sequential data. Unlike traditional feed-forward Neural Networks, RNNs possess a unique architecture that allows them to maintain a memory of past inputs, making them particularly well-suited for tasks such as time series prediction, language modeling, and speech recognition. DuerQues[1], a CNN-based intelligent system was introduced to achieve skill-oriented automatic interview question generation and recommendation. The system paved the way to check such a possibility of textual data in RNN Models. It targets and exploits skill-related knowledge from the usergenerated con tent of online knowledge-sharing communities (KSCs) and the click-through behaviors of search engine queries. Built to work in large amounts of data, RNN is the recurrent connection, which enables the network to retain information about previous inputs and use it to influence the processing of subsequent inputs. This recurrent nature allows RNNs to exhibit dynamic temporal behavior, making them powerful tools for tasks that involve sequences of data points. One key characteristic of RNNs is their ability to handle inputs of varying lengths. This flexibility makes them invaluable in scenarios where the length of the input sequence may vary, such as natural language processing tasks like sentiment analysis, machine
1. INTRODUCTION Generating question papers using Recurrent Neural Networks (RNNs) is a modern approach that leverages advanced technology to streamline the process of exam preparation in educational environments. Despite significant advancements in technology, there remains notable effectiveness in question papers crafted by human experts, which makes the development of RNNs-based automated systems a captivating area of study. In the realm of educational technology, the emergence of CBTs (Computer Based Exams) has set out for a revolutionary, in aspects of convenience, partiality-free marking system, wastage of materials like papers, etc. Recurrent Neural Networks is a deep learning model trained to process and convert a sequential data input into a specific sequential data output. It can generate questions from a sequence of keywords using a sequence-to sequence-based model. Generating question papers using Recurrent Neural Net works (RNNs) is a modern approach that leverages advanced technology to streamline the process of exam preparation in educational environments. The integration of RNNs in question paper generation holds immense potential. These systems not only automate the laborious task of manual question creation but
© 2024, IRJET
|
Impact Factor value: 8.226
|
ISO 9001:2008 Certified Journal
|
Page 1089