Loading article...
Loading article...
Stream/Area:
-
Subject:
-
Paper ID:
IJSREM63657
Volume:
10 |
5 |
May 2026
Pages:
1-3
Publication Frequency:
Monthly (12 issues per year)
License:International Journal of Scientific Research in Engineering and Management follows the Creative Commons Attribution 4.0 (CC BY 4.0) License, allowing free use and distribution with proper attribution.

Computer Science and Engineering
Shri Rawatpura Sarkar University, Raipur, Chhattisgarh, India
Computer Science and Engineering
Shri Rawatpura Sarkar University, Raipur
Legal documents are characterized by dense, jargon-laden language that creates significant comprehension barriers for non-expert users. This paper presents the AI Legal Simplifier (Lexi), a full-stack intelligent web application that automates the analysis, simplification, and explanation of legal documents using Natural Language Processing (NLP), Optical Character Recognition (OCR), and multilingual machine translation. The system accepts PDF, DOCX, and TXT inputs and produces structured plain-language summaries, critical clause extractions, and a quantified risk score (0–100) through a keyword-weighted TF-IDF sentence ranking algorithm. An interactive conversational chatbot enables context-aware document Q&A with multi-turn history and intelligent fallback to OpenAI GPT-3.5. Multilingual output in eight languages — including a custom Hinglish romanisation pipeline — makes legal information accessible to diverse linguistic communities. Additional contributions include server-side TTS/STT voice interface, SHA-256-secured revocable public share links, multilingual ReportLab PDF report generation with Unicode font support, and a side-by-side original-versus-simplified comparison view. The system is engineered on a Flask Blueprint service-layer architecture, uses SQLAlchemy ORM with bcrypt authentication, and is validated by a Hypothesis property-based test suite. Experimental results demonstrate sub-two-second analysis latency, accurate clause and risk detection across diverse legal document types, and production-grade security posture, making Lexi a practically deployable and academically significant contribution to the legal-tech domain.
KEYWORDS
This paper presented the AI Legal Simplifier (Lexi), a comprehensively engineered intelligent web system for legal document analysis. Lexi demonstrates that a practically deployable, secure, and multilingual legal document analysis system can be constructed without depending on large neural models for its primary functionality, using targeted NLP engineering to achieve results competitive with model-based approaches at a fraction of the inference cost. The system's sub-second analysis latency, 0.874 macro-F1 clause detection, 90% risk-score agreement with expert assessments, and production-grade security architecture position it as a significant contribution to democratizing access to legal document intelligence, particularly for multilingual communities in India and beyond
[1] L. Böhme and T. Moore, "The Iterated Weakest Link: A Model of Adaptive Security Investment," IEEE J. Sel. Topics Signal Process., vol. 7, no. 1, pp. 79–88, 2013. [2] J. Devlin, M. Chang, K. Lee, and K. Toutanova, "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding," in Proc. NAACL-HLT, Minneapolis, MN, 2019, pp. 4171–4186. [3] Y. Liu et al., "RoBERTa: A Robustly Optimized BERT Pretraining Approach," arXiv:1907.11692, 2019. [4] I. Chalkidis et al., "LEGAL-BERT: The Muppets Straight Out of Law School," in Findings of EMNLP, 2020, pp. 2898–2904. [5] Y. Koreeda and C. D. Manning, "ContractNLI: A Dataset for Document-Level NLI for Contracts," in Findings of EMNLP, 2021, pp. 1248–1264. [6] H. Zhong et al., "Legal Judgment Prediction via Topological Learning," in Proc. EMNLP, 2018, pp. 3540–3549. [7] C. Grover et al., "Automatic Summarization of Legal Documents," in Proc. Workshop on Legal Text Analysis, Geneva, 2003. [8] V. Sanh et al., "DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter," arXiv:1910.01108, 2019. [9] D. Hendrycks et al., "CUAD: An Expert-Annotated NLP Dataset for Legal Contract Review," in Proc. NeurIPS Datasets & Benchmarks, 2021. [10] A. Conneau et al., "Unsupervised Cross-lingual Representation Learning at Scale," in Proc. ACL, 2020, pp. 8440–8451.
Publication Date
5/21/2026
Peer Review Type
Double-Blind Peer Review
Review Rounds: 2
Documents
Article Views
0
Total views since publication
PDF Downloads
3
Total downloads since publication
Geographic Reach
5
Digital Preservation: This article is permanently archived on Zenodo and complies with all data preservation standards across major indexing platforms.
Plagiarism Screening
This manuscript has been checked for plagiarism using TurnitinPlagiarismChecker software.
Conflict of Interest
The authors declare no conflict of interest related to this research work.
Funding Information
No funding information available for this research.
Related research in the same subject area
Loading similar articles...