Research, experimentation and technical exploration.

Academic and independent work around NLP, retrieval and applied AI, presented with the evaluation context and limitations that matter.

01

BERT-KenMeSH

PythonPyTorchBERTNLP
Research into automatic Medical Subject Headings indexing: a multi-label classification problem for assigning relevant concepts to biomedical articles.
02

Bangla Speech Corpus For Large Vocabulary ASR System

CMUSphinxPocketSphinxHMMSpeech corpus
A speech-corpus and baseline recognition study by Anurag Bhattacharjee and Sujoy Datta, built around standard Bangla as spoken across regions of Bangladesh.