Synthesis of SQL Queries from South African Local Language Narrations

TL;DR

This paper introduces Local-Nar-SQL, a tool that translates South African local language narrations into SQL queries using finite automata, aiming to improve SQL learning for native speakers with limited English proficiency.

Contribution

It presents a novel translation tool leveraging finite automata to convert local language narratives into SQL, enhancing accessibility for native learners.

Findings

Majority of participants found Local-Nar-SQL helpful for understanding SQL.

The tool aids native learners by reducing language barriers in SQL education.

Potential to improve SQL learning outcomes for non-English speakers.

Abstract

English remains the language of choice for database courses and widely used for instruction in nearly all South African universities, and also in many countries. Novice programmers of native origins are mostly taught Structured Query Language (SQL) through English as the medium of instruction. Consequently, this creates a myriad of problems in understanding the syntax of SQL as most native learners are not too proficient in English. This could affect a learner's ability in comprehending SQL syntaxes. To resolve this problem, this work proposes a tool called local language narrations (Local-Nar-SQL) to SQL that uses a type of Finite Machine, such as a Jumping Finite Automaton to translate local language narratives into SQL queries. Further, the generated query extracts information from a sample database and presents output to the learner. This paper is an extension of work originally presented in a previous study in this field. A survey involving 145 participants concluded that the majority found Local-Nar-SQL to be helpful in understanding SQL queries from local languages. If the proposed tool is used as a learning aid, native learners will find it easier to work with SQL, which will eliminate many of the barriers faced with English proficiencies in programming pedagogies.