An Automated Machine Learning NLP Model for Enhancing the Translation of English Causative Sentences into Malayalam Language

This research paper introduces an innovative Neural Machine Translation (NMT) model designed to accurately translate causative English sentences into Malayalam. The study comprehensively explores the various forms of causative sentences utilized in both English and Malayalam, focusing on syntactic and semantic differences. Additionally, it delves into the limitations of current translation models such as Google Translator in accurately translating causative constructs. To address these challenges and enhance translation accuracy, a collaborative effort with professional linguists in both languages led to the creation of a rich dataset. The dataset comprises a diverse collection of causative sentences, covering tense variations, singular and plural forms, noun and verb classes, as well as sentence forms involving non-human and animate-human subjects. This model integrates essential components including tokenization, Part of Speech (POS) tagging, tense identification, suffix identification and addition, and sentence reordering for improved translation. Comparative evaluations between the Google Translator and the newly developed system were conducted based on translation accuracy and fluency. The evaluation revealed that the novel NMT model consistently outperforms existing translation systems, demonstrating superior translation results. The findings underscore the effectiveness and potential of the proposed NMT model in advancing causal English to Malayalam translation, offering valuable insights for the improvement of machine translation systems.