College Lens: A Verified Student Review and Admission Prediction System using OCR and Machine Learning

Abstract

In the digital era, students increasingly depend on online platforms to evaluate colleges based on reviews, placements, and facilities. However, existing systems often allow unverified users to submit feedback, resulting in fake, biased, and misleading information. To address these challenges, this paper proposes College Lens, a web-based platform that ensures authenticity and transparency through verified student reviews and intelligent admission prediction. The system utilizes Optical Character Recognition (OCR) to validate user authenticity by extracting and verifying information from academic documents such as mark sheets and fee receipts. Only verified students are permitted to post reviews, ensuring reliability. Additionally, a machine learning model built using Scikit-learn predicts suitable colleges based on student academic performance, including CET scores and 12th-grade percentages. The platform is developed using React for the frontend, Django for backend processing, and MySQL for efficient data management. By integrating verification mechanisms with predictive analytics, the proposed system provides a trustworthy and data-driven environment that assists students in making informed academic decisions.

Introduction

Currently, most college review websites suffer from fake reviews, lack of verification, and biased opinions, which mislead students. They also do not help students predict their admission chances based on academic performance.

To solve this, College Lens introduces a system where only verified students can submit reviews. Verification is done using OCR (Optical Character Recognition) to check documents like mark sheets and fee receipts, ensuring authenticity. The platform also includes a machine learning model that predicts suitable colleges based on student data such as marks and entrance scores.

A literature review shows that earlier systems focused on recommendations and reviews but failed to properly verify users, leading to unreliable results. The proposed system improves this by combining document-based verification and predictive analytics.

The main objectives include ensuring authentic reviews, removing fake feedback, helping students choose colleges using data-driven predictions, improving transparency, and providing centralized college information such as cutoffs, placements, and facilities.

In terms of methodology, students register, upload documents for OCR verification, and after approval can post reviews. Colleges also provide official data. All information is stored in a MySQL database, and an ML model predicts suitable colleges based on student input. Admins monitor the system to maintain accuracy and prevent misuse.

The system uses a three-tier architecture:

• Frontend: React (user interface)
• Backend: Django (logic and APIs)
• Database: MySQL (data storage)
  with additional OCR and ML modules.

The results show that OCR verification effectively reduces fake reviews, and the ML model achieves about 90–95% accuracy in recommending colleges. Overall, the system improves transparency, trust, and decision-making for students by combining verified feedback with intelligent predictions.

Conclusion

The proposed system, College Lens, effectively addresses the limitations of existing college review platforms by introducing a verified and intelligent solution. By combining OCR-based authentication with machine learning-based prediction, the system ensures both reliability and accuracy in student decision-making. The platform enhances transparency, eliminates fake reviews, and provides personalised college recommendations. Future enhancements may include sentiment analysis, mobile application development, and integration with official admission portals to further improve system efficiency and user experience.

References

[1] R. Sharma and A. Gupta, “Enhancing College Review Systems Using Verified Student Data and Machine Learning,” International Journal of Advanced Research in Computer Science and Software Engineering, vol. 15, no. 3, pp. 45–52, 2025. [2] S. Patel and P. Mehta, “AI-Based Student Verification for Authentic Academic Reviews,” Journal of Artificial Intelligence and Data Science, vol. 12, no. 1, pp. 120–128, 2024. [3] V. Rao and K. Singh, “Building Trust in Educational Platforms Through Document-Based Verification,” IEEE Transactions on Learning Technologies, vol. 17, no. 2, pp. 88–95, 2024. [4] D. Kumar and R. Nair, “Predictive Analytics in College Admissions: A Machine Learning Approach,” Journal of Educational Data Mining, vol. 14, no. 4, pp. 67–75, 2023. [5] M. Thakur and J. Bansal, “Reducing Fake Reviews in Student Feedback Systems Through Verified Authentication Models,” in Proc. International Conference on Intelligent Computing and Data Engineering (ICICDE), pp. 58–65, 2023. [6] S. Deshmukh and P. Kaur, “Review Authenticity in Higher Education: A Framework for Verified Student Reviews,” Journal of Emerging Technologies in Web Intelligence, vol. 9, no. 3, pp. 134–142, 2022. [7] N. Alam and S. Roy, “Integrating Django and React for Scalable Educational Review Portals,” International Journal of Computer Applications, vol. 184, no. 5, pp. 99–106, 2022. [8] T. Bose and M. Iyer, “Data-Driven Insights for College Admissions Using Student Performance Metrics,” Procedia Computer Science, vol. 192, pp. 411–418, 2021. [9] L. Fernandez and A. Das, “A Machine Learning Framework for Transparent and Reliable Student Feedback Systems,” International Journal of Information and Education Technology, vol. 10, no. 9, pp. 710–717, 2020.

Copyright

Copyright © 2026 Mr. Shreyash S. Gopale, Ms. Shraddha D. Deshmukh, Ms. Muskan V. Tandel, Mr. Arjun S. Jondhale, Mr. Sharad M. Rokade. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.