VoxDB: Voice Command Based Data Retrieval System

Voice-based database interaction has gained promi- nence as an intuitive and efficient method for retrieving informa- tion. This paper presents a literature survey on speech-to-SQL query generation, emphasis on a rule-based approach imple- mented in our project. While machine-learning-based techniques , they often require extensive training data and may struggle with adaptability. Rule-based approach can ensures greater accuracy and consistency by explicitly mapping natural language components to structured SQL queries. For real-time speech-to- text conversion Our system integrates webkitSpeechRecognition API, followed by Natural Language Processing (NLP) techniques for query generation. The system processes the input text by manually removing stopwords using a predefined list of stop- words. The extracted entities and conditions from the input are then structured into SQL queries based on predefined database schema rules. This survey examines existing research in speech recognition, NLP-based text processing, and rule-based query generation, highlighting their advantages and limitations. Our findings suggest that especially for domain-specific databases a hybrid approach combining rule-based methods with NLP techniques significantly improves SQL query accuracy. The structured framework enhances interpretability, adaptability, and efficiency, making it a reliable alternative to purely machine- learning-driven models.This paper emphasizing the effectiveness of a rule-based approach in ensuring precise and structured SQL query formation also provides insights into the method- ologies, challenges, and advancements in speech-driven database retrieval.