![References
Textbooks and Journals:
•AI grading platforms and OCR-based systems are also discussed in various academic books and journals. Some platforms like Akindi and
Crowdmark are explored for their ability to automate the grading process in a structured way, which directly applies to both image recognition
and feedback in grading.
•Relevant research can be found in AI-focused journals like Journal of Artificial Intelligence Research (JAIR) and IEEE Transactions on
Learning Technologies.
List of papers/books/websites etc. refer for project
Online journals with a DOI [Digital Object Identifier]
Bao, X. A., Dai, S. C., Zhang, N., Yu, C. H, International Journal of Grid and Distributed Computing, DOI
Text Similarity Based on Graph Connectivity:-
http://article.nadiapub.com/IJGDC/vol9_no4/9.pdf DOI](https://image.slidesharecdn.com/autograderautosaved-2-250803113705-e4165040/85/Automatic-Grading-of-marksheet-engineering-ppt-14-320.jpg)
Automatic Grading of marksheet engineering ppt
- 1. Guide: Incharge: Prof. Ashish Mishra Prof. Manish Motghare Assistant professor Assistant Professor Name of Projectees -Gaurav Khandedia -Harshal Sarve - Harshal Rode - Himanshu Mankar Session 2024-2025 AI tool for Automatic Grading of Marksheet Introductory Project Seminar Of Minor Project under Course Category Community Engagement Project (CEP) On G H RAISONI COLLEGE OF ENGINEERING AND MANAGEMENT (Approved by AICTE, New Delhi and Recognized by DTE, Maharashtra) An Autonomous Institute Affiliated to Rashtrasant Tukadoji Maharaj Nagpur University, Nagpur Accredited by NAAC with A+ Grade Department of Artifical Intelligence
- 2. • Introduction • Impact of project on environment • Literature Survey • Problem Statement • Justification for selecting the Title • Objective of the Project • Design/Block Diagram/Circuit Diagram • Description of the Project Working • Result and Discussion • Conclusion and Future Scope • References Contents
- 3. Introduction • AI tool for automatic grading of answer sheets uses Machine Learning and Natural Language Processing to evaluate and grade student answers. • The tool analyzes each answer and compares it with the expected answer provide by the teacher to assess correctness, relevance, and completeness. • It saves teachers time, identifies areas of weakness, and offers personalized feedback for improvement. • Advanced technologies like Optical Character Recognition (OCR) and string similarity algorithms, the system compares student responses with the teacher's correct answers to determine a score. • Automatic grading systems can handle various assessment formats, including multiple-choice questions, short answers, and even essays. • This technology not only enhances the learning experience but also allows teachers to focus more on instruction and student engagement. • These tools leverage machine learning algorithms and natural language processing to evaluate student performance efficiently and accurately. AI Tool For Automatic Grading Of Answer Sheet
- 4. Impact of project on environment Impact of AI Tool for Automatic Grading of Marksheets Environment: • Less Paper: Using AI for grading can reduce the need for paper, which is better for the environment. • Reduced Energy Use: Computers can be more energy efficient than humans, especially when grading a large number of papers. Society: • Faster Results: Students get their grades quicker, which can help them understand their progress and improve. • Fairness: AI can grade more fairly, without human bias. • More Time for Teachers: Teachers can spend more time helping students and planning lessons.
- 5. Literature Survey Sr.no. Paper Title and its Author Details of Publication Findings 1. A Novel Implementation of Marksheet Parser Using Paddle-OCR Authors:- Sankalp Bageria, S Irene, Harikrishna, Elakia V M Centre for Development of Advanced Computing, Chennai, India Working of an OCR 2. AI Assistance Aid in the Grading of Handwritten Answer Sheets Authors:- Pritam Sil, Parag Chaudhuri, Bhaskaran Raman IIT Bombay This work introduces an AI- assisted grading pipeline 3. Automatic Program Assessment, Grading and Code Generation: Possible AI-Support in a Software Development Course Authors :-Uwe M. Borghoff, Mark Minas, Kim Mönch University of the Bundeswehr Munich Institute for Software Technology Department of Computer Science Automatic grading with the support of AI 4. A Machine Learning Approach for Automated Evaluation of Short Answers Using Text Similarity Based on WordNet Graphs. Author:- Atoum I and otum A Wireless personal communication . This research focuses on evaluating short text answers using machine learning techniques, particularly with text similarity based on WordNet.
- 6. Problem Statement Problem identified and its justification Concerns about the traditional grading system as time-intensive, subjective and error-prone would not persist. This creates a place a large burden on educators, taking up time they could spend doing other necessary tasks such as coaching personalized learning engagement. • Time Constraints :- Time is a vast barrier for educators, forces, constraints that made it impossible toto provide personalized feedback and communicate individually student needs. • Subjectivity Bias:- Human grading can be subjective, leading to inconsistent evaluations and unfair assessments. • Grading Errors:- But the weakness of manual grading human errors, potentially impacting student performance and motivation. • Delayed Feedback to Students: Traditional grading methods can lead to slow feedback delivery, making it difficult for students to learn from their mistakes in a timely manner. Without quick feedback, students may not be able to improve before their next assessment.
- 7. Justifications for Selecting the Title Justifications for Selecting the Title • Traditional grading methods are labor-intensive and time-consuming, especially for large classes. • Sometimes teachers or faculties are confused in true or false due to large quantity of questions present in paper, so sometimes mistake are happened while checking the true or false which are solved by students during exam. • Short or long answer are most challenging part for teachers while checking the large group of answer sheet, and we know that toady's generation student making their own answer which are related to topic but in their own understanding language. • With the integration of OCR and string similarity algorithms, the system accurately compares student responses to the teacher’s answers., using OCR technology to automatically grade the exams. • The system is designed to work with images of handwritten answer sheets. It makes grading more versatile by handling different formats, such as printed or scanned documents, making it adaptable to a variety of exam types and educational contexts.
- 8. Objective are as follow:- Objectives of the Project • Efficient grading process: It evaluates the quality of a students work and it also organizes to mark transitions. • Consistent evaluation standards : It ensures fair , objective and reliable assessment by applying uniform criteria across all evaluation. • Reduce manual workload : It involves automating tasks to save time, increase efficiency and minimize human effort. • Reliability : It involves consistency and accuracy of evaluation results when the same standards are applied repeatedly. • Provide detail feedback : It means offering specific ,clear and constructive information or comments about someones performance.
- 9. Design/Block Diagram/Circuit Diagram with description START LOAD PAGE USER INTERACTS UPLOAAD TEACHER’S ANSWER PAGE UPLOAD STUDENT ANSWER PAGE GRADE BUTTON PROCESSING FILE HANDLING (UPLOAD/READ) IMAGE RECOGNITION (TESERACT.JS) TEXT COMPARISON (STRING SIMILARITY) OUTPUT DISPLAY Grade (accuracy and result (Parameters ) 1. similarity > 0.9 grade ='A+' 2. similarity > 0.8 grade = 'A' 3. similarity > 0.6 grade = 'B' 4. similarity > 0.4 grade = 'C 5. similarity > 0.2 grade = 'D’ 6. similarity > 0.1 grade = 'E’ 7. similarity>0.0 grade=fail Store the output in history section
- 10. Description of the project working Processing:- •File Handling • File Selection • File Reading •Image Recognition • Tesseract.js for OCR •Text Comparison • Use String Similarity to compare text User Interface (UI):- •Upload Section • Upload Teacher's Answer Pages • Upload Student's Answer Pages •Preview Section • Display Teacher's Answer Previews • Display Student's Answer Previews •Grade Button •Result Display Output:- •Final grading results displayed to the user PARAMETER OF GRADING:- 1. similarity > 0.9 grade ='A+' 2. similarity > 0.8 grade = 'A' 3. similarity > 0.6 grade = 'B' 4. similarity > 0.4 grade = 'C 5. similarity > 0.2 grade = 'D' 6. similarity > 0.1 grade = 'E' 7. similarity>0.0 grade=fail
- 11. Result & Discussion Result & Discussion of the project Result • Increased Efficiency Educators will be able to save time and focus on more engaging teaching methods. • Enhanced Accuracy Consistent and grading will ensure fairness and accurate evaluation. • The system is scalable and can handle grading for a large number of students, making it suitable for institutions.. • Personalized Learning Tailored feedback will empower students to take ownership of their learning and improve their understanding. • The use of OCR technology ensures accurate recognition of handwritten or printed answers. Discussion: • Limitations: The tool may struggle with very messy handwriting or unusual grading systems. • Human Oversight: Human teachers should still review the AI's grades to catch any errors. • Potential for Improvement: The tool can continue to improve with more data and advancements in AI technology.
- 12. Conclusion & Future Scope Conclusion: • AI grading tools have the potential to revolutionize education by making grading more efficient, fair, and accurate. • These tools can save time for teachers, provide students with faster feedback, and reduce the risk of human error. • Efficiency: AI grading tools can save teachers a lot of time by automating the grading process. • Fairness: These tools can help ensure that grading is more consistent and objective, reducing bias. Future Scope: • Customization: AI tools can be further customized to meet the specific needs of different schools and grading systems. • Integration: Future tools may be able to integrate with other educational software, providing a more comprehensive solution. • Customization: AI grading tools can be made to fit the specific needs of different schools and grading systems. • Integration: Future tools might work well with other educational software, making it easier to use. In conclusion, AI grading tools offer a promising solution for the future of education. As technology continues to advance, we can expect to see even more innovative and effective tools that can benefit both students and teachers.
- 13. References List of papers/books/websites etc refer for project Online journals A Novel Implementation of Marksheet Parser Using Paddle-OCR:- https://www.researchgate.net/publication/382331465_A_Novel_Implementation_of_Marksheet_Parser_Using_PaddleOCR Google Scholars Yang, X., Huang, Y., Zhuang, F., Zhang, L., Yu, S.: Automatic Chinese short answer grading with deep autoencoder. In: Penstein Rosé, C., et al. (eds.) AIED 2018. LNCS (LNAI), vol. 10948, pp. 399–404. Springer, Cham (2018). https://doi.org/10.1007/978-3-319-93846-2_75 Online journals AI Assistance Aid in the Grading of Handwritten Answer Sheets:- https://www.researchgate.net/publication/381893086_Can_AI_Assistance_Aid_in_the_Grading_of_Handwritten_Answer_Sheets Automatic Program Assessment, Grading and Code Generation: Possible AI-Support in a Software Development Course
- 14. References Textbooks and Journals: •AI grading platforms and OCR-based systems are also discussed in various academic books and journals. Some platforms like Akindi and Crowdmark are explored for their ability to automate the grading process in a structured way, which directly applies to both image recognition and feedback in grading. •Relevant research can be found in AI-focused journals like Journal of Artificial Intelligence Research (JAIR) and IEEE Transactions on Learning Technologies. List of papers/books/websites etc. refer for project Online journals with a DOI [Digital Object Identifier] Bao, X. A., Dai, S. C., Zhang, N., Yu, C. H, International Journal of Grid and Distributed Computing, DOI Text Similarity Based on Graph Connectivity:- http://article.nadiapub.com/IJGDC/vol9_no4/9.pdf DOI
- 15. • Burstein, J., Baker, R. S., & Kim, Y. (2013). Automated essay scoring: A survey of the state of the art. Journal of Educational Measurement, 50(4), 389-414. • Chen, Y., Wang, H., & Li, B. (2019). A deep learning approach for automated essay scoring. Computers & Education, 137, 103-113. • Kemp, C., Nguyen, H., & Chan, C. (2017). Ahybrid approach to automated essay scoring. Computers & Education, 112, 1-13. • Li, H., Zhang, Y., & Liu, Q. (2021). A transformer- b ased approach for automated essay scoring. Computers & Education, 169, 104220. • Liu, Y., Wang, Z., & Zhang, Q. (2020). A deep learning approach for automated multiple-choice question grading. Computers & Education, 155, 103796. • Wang, H., Chen, Y., & Li, B. (2022). A deeplearning approach for automated short answer question grading. Computers & Education, 183, 104137. • Wang, H., Chen, Y., & Li, B. (2023). A deep learning approach for automated essay scoring based on semantic similarity. Computers & Education, 196, 104471. • Automated Scoring for the Assessment of Common Core Standards. Retrieved from http:// research.collegeboard.org/sites/default/files/ publications/2012/8/ccss-2010-5-automated- scoring-assessment-common-core-standards.pdf • A Framework for Evaluation and Use of Automated Scoring. Educational Measurement: Issues and Practice, 31, 2–13. • Investigating the judgmental marking process: an overview of our recent research. Research Matters, 5, 6–9. • Mental model comparison of automated and human scoring. Journal of Educational Measurement, 36, 158–184. References
- 16. Educational Testing Service (ETS): • ETS Essay Grader. https://www.ets.org/Media/ Research/pdf/RD_Connections_21.pdf • Automated Essay Scoring (AES) Toolkit. https:// dl.acm.org/doi/abs/10.1145/3493700.3493765 • https://www.unesco.org/en/digital- education/artificial-intelligence • An evaluation of IntelliMetric™ essay scoring system using responses to GMAT® AWA prompts. • Retrieved from http://www.gmac.com/~/media/ Files/gmac/Research/research-report-series/ References
- 17. Thank you ! Nagpur | Pune | Jalgaon | Amravati | Pandhurna | Bhandara