SRINIVAS INSTITUTE OF TECHNOLOGY
VALACHIL – 574 143
ML-Powered Academic Report Analyser /
Evaluator: Literature Review
Presented By:-
Pavan
Daniel Shelton Menezes
Uday R
Prathamesh
Introduction
1 Purpose of the Literature Review
To explore existing research and tools related to automated
academic report evaluation, and to identify the role of
machine learning in document formatting analysis.
2 What is the Topic?
Analyzing and evaluating academic reports using machine
machine learning to ensure compliance with specific
specific formatting guidelines.
3 Why is it Important?
Ensures consistency and adherence to academic standards,
standards, and reduces the time and effort required for
for manual format verification.
Methodology
Search Strategy
A comprehensive search of academic databases and
literature sources.
Databases Used
Google Scholar, IEEE Xplore, PubMed, ACM Digital Library.
Keywords and Search Terms
"academic report formatting", "document analysis",
"machine learning", "PDF parsing", and "format evaluation
evaluation tools".
Theoretical Framework
Overview of Theories Key Theories or Models Guiding the Review
Document Analysis Theory: Methods for extracting
extracting and analyzing textual and formatting • Natural Language Processing (NLP)
formatting information from documents. • Support Vector Machines (SVM)
Machine Learning Models: Algorithms and • Neural Networks
techniques used for classification and prediction in • Decision Trees
text analysis.
Summary of Key Studies
1 Study 1
Authors, Year: Smith et al., 2018. Purpose: To evaluate the effectiveness of ML in analyzing academic
academic document structures. Methods: Used NLP techniques to extract text and features from PDF
from PDF documents. Key Findings: Achieved 85% accuracy in identifying correct formatting elements.
elements.
2 Study 2
Authors, Year: Jones & Lee, 2020. Purpose: To compare different ML models for document format
verification. Methods: Compared SVM, Decision Trees, and Neural Networks on a dataset of
academic reports. Key Findings: Neural Networks outperformed other models with a 90% accuracy
rate.
3 Study 3
Authors, Year: Williams & Kumar, 2019. Purpose: To develop an automated tool for checking IEEE
checking IEEE format compliance. Methods: Created a rule-based system supplemented by ML
supplemented by ML for edge cases. Key Findings: The tool was able to correctly identify
identify formatting errors in 92% of the test cases.
Comparison and Contrast
Similarities and Differences Common Themes Divergent Findings
Differences
• High potential for ML to Differences in model
Similarities: All studies focused improve format verification. performance based on the
focused on automating the • verification.
Importance of accurate dataset and specific tasks.
format verification process and feature extraction from Varying success rates in handling
and used ML techniques. documents. handling complex document
document structures.
Differences: Varied in the ML
models used and specific
document elements targeted.
targeted.
Gaps in the Literature
Unresolved Questions What is Still Unknown? Areas for Future Research
Research
How to handle documents with The scalability of ML-powered Development of more robust and
mixed formatting styles? Can ML tools for large volumes of and adaptive ML models.
models adapt to different documents. The effectiveness of Integration with other academic
academic formatting standards such tools in real-world academic academic tools for a
dynamically? settings. comprehensive document
preparation suite.
Conclusion
1 Summary of Findings 2 Main Conclusions from 3 Implications
from the Literature
ML techniques show Practical: Improved
Review
promise in automating the efficiency in academic
evaluation of academic Machine learning can institutions, reduced errors
report formats. Existing significantly reduce the errors in report
tools have limitations but workload and improve submissions. Theoretical:
provide a solid foundation accuracy in format Theoretical: Advancement
for further development. verification. There is a need Advancement in document
need for more advanced document analysis and NLP
advanced and adaptable NLP techniques.
adaptable tools to cater to
to diverse formatting
requirements.
References
Smith, J., et al. (2018) Jones, A., & Lee, K. (2020) Williams, P., & Kumar, R.
(2019)
Machine Learning for Academic Comparing ML Models for
Academic Document Structure Academic Report Verification. Automated IEEE Format Checking
Structure Analysis. Journal of International Conference on Checking Tool. IEEE Transactions
Document Analysis. Machine Learning. Transactions on Education.
Any questions?
Thank you