Machine Reading Comprehension Using NLP

Machine Reading Comprehension (MRC) is a type of question-answering mechanism that involves reading documents and responding to questions. The challenge of enabling machines to effectively interpret paragraphs, comprehend related inquiries, and generate appropriate responses has long persisted. Unlike humans, who possess the ability to grasp the logic and intent behind a question and formulate a fitting reply, machines struggle with this task. Thus, the primary goal of this project is to leverage Natural Language Processing (NLP) to enable machines to read paragraphs, understand questions, and provide coherent answers. The input for the reading comprehension model is the question and the context. The passage’s answer is the model’s output. A sort of writing and reading comprehension (RC) is becoming more and more popular in the study of natural language processing (NLP). Datasets like SQUAD (Stanford question answer dataset) are used. Various models like BERT, and ELECTRA, are used to predict the answers to respective questions. There are a few models that can answer the questions in the Telugu language. In this analysis, the MRC is implemented in the Telugu language using BERT model. Although existing MRC models surpass humans on numerous benchmark datasets in English language comprehension, there remains a significant disparity between these models and real-world human-level reading comprehension.