Copyright © 2026 Authors retain the copyright of this article. This article 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.

@article{195787,
        author = {Abhishek Ram and Aman Sahani and Atharv Limje and Dr. Sarita Patil},
        title = {ANALYZING LEGAL TEXTS USING LEGAL LONGFORMER FOR DOCUMENT SUMMARIZATION},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {11},
        pages = {2558-2566},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=195787},
        abstract = {Reviewing lengthy and complex legal documents is a predominantly manual, time-consuming process that is highly susceptible to human error. To address this bottleneck, this paper presents an automated, end-to-end Natural Language Processing (NLP) pipeline designed to streamline the extraction, semantic search, and summarization of legal contracts. The proposed architecture ingests digital PDFs using PyMuPDF and processes the unstructured text through a sequence of domain-specific language models. Specifically, Legal-ALBERT, evaluated against the Contract Understanding Atticus Dataset (CUAD), is utilized to accurately classify and extract critical legal clauses. Concurrently, Legal-Longformer generates dense document embeddings to enable context-aware semantic search, bypassing traditional keyword limitations. To synthesize the retrieved information, Long-T5 performs abstractive summarization, distilling multi- page documents into concise overviews. Furthermore, a translation layer is integrated to support multilingual accessibility. Evaluated using F1, ROUGE, and BLEU metrics, the system demonstrates high accuracy and consistency, effectively reducing document review time from hours to minutes while successfully handling documents up to 100 pages in length.},
        keywords = {LegalTech, Natural Language Processing, Abstractive Summarization, Semantic Search, Legal-ALBERT.},
        month = {April},
        }