arXiv:2606.00116v1 Announce Type: new Abstract: This study introduces a novel architecture of KAN-based BiGRU model for the task of classification and summarization of legal documents in a low-resource multilingual setup. In order to tackle problems associated with domain language, the usage of different languages, long dependencies within context, and class imbalance, we employ the dataset composed of legal documents from Bangladesh and taken from Manupatra, which include Bengali, English, and transliterated Bengali languages. Our classification task involves BiGRU model, along with Kolmogoro
The proliferation of AI models necessitates specialized approaches for low-resource languages and domain-specific challenges, driving innovation in fine-tuned architectures like KAN-based BiGRU.
This research addresses the critical need for effective AI in legal domains and non-English languages, which are often underserved by mainstream models, unlocking new markets and improving access to legal tech.
The ability to accurately classify and summarize legal documents in multilingual, low-resource contexts improves, potentially lowering barriers to legal information access and automating legal workflows.
- · Legal Tech Developers
- · Law Firms in developing nations
- · NLP researchers
- · Judiciaries in multilingual regions
- · Translators for legal documents
- · Manual legal document review processes
Improved efficiency in legal services and research, particularly in countries with diverse linguistic landscapes.
Expansion of AI-powered legal services into new geographical and linguistic markets.
Enhanced legal transparency and access to justice due to automated understanding and summarization of complex legal texts.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.CL