editorinchief@ijsreat.com +91 9840521421 Scholarly Open Access, Peer-reviewed, Refereed Journal e-ISSN 2583-1240 Since 2021
  1. Home
  2. Archives
  3. 2026
  4. Vol. 6 Issue 6
  5. Article
Original Article

Pariksha AI: An Intelligent Curriculum-Grounded Question Paper Generation System Using Generative Artificial Intelligence

  • Paper ID 435
  • Volume 6, Issue 6
  • Pages 196-202
  • Published Online: November-December 2026
Open Access Peer Reviewed Crossref DOI

Authors

  • Dr. Naveen H M1
  • Sanjana Gowda2
  • Rakshitha D S3
  • Vinayaka P J4
  • Prajwal S5

1 Associate Professor, Department of Computer Science and Engineering, Bapuji Institute of Engineering and Technology, Davangere, Karnataka, India. 2 3 4 5 Department of Computer Science and Engineering, Bapuji Institute of Engineering and Technology, Davangere, Karnataka, India.

Article Information

Published Online
November-December 2026
Pages
196-202
Article Type
Original Article
Check for updates
check update

Abstract

View PDF

Question-paper preparation is a recurring academic task that requires more than writing a set of questions. A teacher must interpret the prescribed curriculum, decide the distribution of marks, select suitable question types, maintain an appropriate level of difficulty, and ensure that the paper reflects intended learning outcomes. This paper presents Pariksha AI, a web-based system that combines curriculum retrieval with a Large Language Model (LLM) to assist teachers in preparing structured question papers. The implemented application accepts curriculum material in PDF or text form, extracts its textual content, divides the material into overlapping chunks, and ranks relevant chunks using terms derived from the selected subject and chapters. The retrieved curriculum context is combined with a teacher-defined examination blueprint containing board, grade, subject, chapters, total marks, duration, difficulty, language, question counts, and additional instructions. The resulting prompt is submitted through the OpenRouter API and the model is instructed to return strict JSON containing questions, answers, marks, difficulty and Revised Bloom’s Taxonomy labels. A deterministic validation layer checks the total number of questions, required question and answer fields, and exact total marks before the paper is accepted. Django authentication provides teacher-specific ownership of uploaded documents and generated papers, while ReportLab is used to create separate question-paper and answer-key PDFs. The system is intended as a human-in-the-loop assessment authoring assistant: generative AI drafts the content, software enforces measurable constraints, and the teacher retains responsibility for final academic review.

APC from ₹1200 + 18% GST

Peer-reviewed open access₹1200 Indian (₹1400 with DOI)$80 USD internationalpaid after acceptance Peer-reviewed open access₹1200 Indian$80 USD international ₹1200 Indian · $80 intl

  • CrossRef DOI
  • ISSN 2583-1240
  • Open Access
  • Peer Reviewed
  • Plagiarism Checked
  • Google Scholar
Submit Paper