banglabert

This repository contains the official release of the model "BanglaBERT" and associated downstream finetuning code and datasets introduced in the paper titled "BanglaBERT: Language Model Pretraining and Benchmarks for Low-Resource Language Understanding Evaluation in Bangla" accpeted in Findings of the Annual Conference of the North American Chap…

This research examines the performance of Large Language Models (GPT-3.5 Turbo and Gemini 1.5 Pro) in Bengali Natural Language Inference, comparing them with state-of-the-art models using the XNLI dataset. It explores zero-shot and few-shot scenarios to evaluate their efficacy in low-resource settings.

Enhancinng Bengali Text Summarization via Ranking based Approach

Bengali Hate Speech Detection Using BanglaBERT and XLM-R : This repository showcases an advanced implementation for detecting Bengali hate speech using **BanglaBERT** and **XLM-R** models. It incorporates cutting-edge deep learning techniques, preprocessing methods, and tools to classify Bengali text into hate or Non_hate.

Automatic Punctuation Prediction Using BanglaBERT

Bengali Misogyny Identification with Deep Learning and LIME.

Enhancinng Bengali Text Summarization via Ranking based Approach

Repository for my undergraduate thesis

First Bangla Language Processing Workshop @ EMNLP 2023: Bangla Sentiment Analysis

Official IEEE SPICSCON 2025 research repository: BanglaASTE framework for Aspect-Sentiment-Opinion Triplet Extraction in Bangla e-commerce using BanglaBERT and XGBoost (89.9% Acc / 89.1% F1).

Official implementation of "BanglaSentNet: A Hybrid Deep Learning Framework for Multi-Aspect Sentiment Analysis in Bangla E-Commerce Reviews" (ICDSAIA 2025, Springer CCIS vol. 2682).

Official repo for "BanglaMM-Disaster" (IEEE SPICSCON 2025). Multimodal deep learning framework fusing Transformers (mBERT, BanglaBERT, XLM-R) and CNNs (ResNet50) for 9-class disaster classification in Bangla social media. Achieves 83.76% accuracy (+16.91% over vision, +3.84% over text).

Presentation slides, materials, and certificate from 3MT Bangladesh 2025 — based on our published IEEE SPICSCON 2025 paper on multimodal disaster classification.