Machine Learning Techniques to Understand Partial and Implied Data Values for Conversion of Natural Language to SQL Queries on HPCC Systems

Abstract

There has been an exponential growth in the amount of data produced daily in recent years, owing to the widespread use of technology. Ease of access to this data is of utmost importance in this day and age. Although in the past, use of structured query languages to query the data stored in the hard-drive was satisfactory, use of natural language to access the data is more desired. This paper talks about mapping partial data values to its corresponding data values and attributes in the schema to enrich the natural language query. Machine learning algorithm, Long Short Term Memory, preceded by an Embedding layer has been used on the HPCC Systems platform. The resulting model gives an accuracy of 99.6%, while its implementation with the experimental setup gives an accuracy of 92%.