Advancing deep learning techniques for low-resource Shahmukhi Punjabi language processing
Muhammad Shabbir, Mudassir Iftikhar · International Journal of Complexity in Applied Science and Technology · 2026
Shahmukhi Punjabi, an underrepresented language in computer linguistics, is being processed in a variety of ways thanks to the growing interest in natural language processing (NLP). This paper explores the software of Named Entity Recognition (NER), Recurrent Neural Network (RNN), and Long Short-Term Memory (LSTM) models on a dedicated Shahmukhi Punjabi dataset. The study includes a thorough analysis of loss graphs, accuracy measures, and misunderstanding matrices, providing insightful information on how well certain designs operate. Recent years have seen a tremendous rise in the discipline of Natural Language Processing (NLP), with a growing emphasis on languages that have historically been underrepresented in computational linguistics. This research looks into Shahmukhi Punjabi's underappreciated work on a variety of sophisticated complex space models, including named entity recognition (NER) and network (RNN). Shahmukhi Punjabi, a language spoken primarily in Pakistan, presents particular difficulties for electronic linguistics because of its unusual spelling and subtle linguistics. The majority of current research focuses on languages that are widely spoken, leaving a gap in our knowledge of computer techniques comparable to Shahmukhi Punjabi. By implementing LSTM, RNN, and NER models and evaluating their efficacy on specific Shahmukhi Punjabi data, this work seeks to close this gap. To provide a more comprehensive view of model performance, the study looks at confusion matrices and loss graphs in addition to standard accuracy measurements. Our LSTM Model Give us The 82% Accuracy And RNN give US The 82.57%.