A Corpus Based SQL Formation From Bangla Language Using Neural Machine Translation

Natural Language Processing (NLP) is a field of study that focuses on enabling computers to process, understand, and generate human language. With the rapid growth of digital data, NLP has become increasingly important in recent years. One of the subfields of Natural Language Processing (NLP) is Neural Machine Translation (NMT) which deals with the automatic translation of one language into another using artificial neural networks. One of the main advantages of NMT is that it can capture the context of a sentence better than traditional Statistical Machine Translation (SMT) systems. These advancements in Natural Language Processing enables us to form SQL query from natural language query. Although work is present in the English Language, no research has been conducted to generate SQL queries from the Bangla language. To fill the gap, our work provides guidelines and steps to generate SQL queries from the transliterated version of the Bangla language. Considering one scenario of the hospital database, 2844 text queries and relevant SQL queries were formed to train the NMT model. The highest accuracy achieved to predict the SQL query from the model was 99.85 percent.