BDeedNet: A Deep Learning Framework for Bengali Deed Summarization
Bibliographic record
Abstract
Legal documents, also known as deeds, are considered confidential documents for every individual. However, many find it a daunting task to read and navigate through these documents, which often contain a large number of pages. Researchers worldwide are striving to implement models and adopt strategies to address these challenges in the AI era. This study proposes a novel solution to these issues by summarizing Bengali deeds using a deep learning approach. This article employed transformer-based models with Google mT5 and Bangla T5 algorithms to evaluate human-annotated summaries using standard NLP metrics such as BLEU, ROUGE, WIL, WER, and METEOR. This study also introduces "BDeedNet," a specialised dataset created for summarising Bengali legal deed texts, addressing the lack of resources in this area, which is the first of its kind and enabling more precise and meaningful summaries of legal texts on the proposed models. Each model performs well on the BDeedNet dataset with high accuracy and efficacy, with Bangla T5 outperforming other models and yielding the best results, demonstrating significant potential for real-world applications in legal document summarization. This study sets the path for future developments in Bengali legal text processing and adds significant tools for legal and linguistic communities.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.005 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".