Automated essay scoring system for Bangla language

Abstract

A plethora of scholarly articles have been thoroughly analyzed, scrutinized and delved into to decipher the growing need of Automated Essay Scoring (AES) system for Bangla and its implications on the field of education, language assessment, publishing industries, language learning platforms and other multiple distinct do mains. Via a fusion of global vantage points and melding a wider international outlook with a detailed inspection on Bangladesh's local landscape, the study has identified distinct nuances and discerned correlations pertinent to the context of the nation regarding the repercussion and significance of the AES. The pressing necessity and demand for the implementation and integration of AES within the sectors of Bangladesh is palpable and exceedingly essential accentuated by the lacuna in this very critical domain. This study adeptly bridged the gap between the international trends and the intricacies of the automated essay scoring for Bangla furnishing the localized perceptions on the shifting dynamics of language assessment. Harnessing the potential of the growing Natural Language Processing (NLP) domain and making use of its powerful neural network approaches and all its alterations including the traditional transformer models like BERT and the advanced large language models (LLMs), our primary aim had been to effectively focus on these vacancies within the language assessment techniques and reshape it by modeling an automated essay scoring system for Bangla which can accurately assess and evaluate the Bangla based essays countering any biases by being completely transparent establishing a fair assessment of the dissertations and the written scripts with a commitment to elevate student learning, empowering educators and propelling advancements in educational ramifications.