An Arabic intelligent interface for accessing NoSQL document-oriented databases

NoSQL document-oriented databases are widely used to manage large volumes of unstructured and semi-structured data, especially in web applications and Big Data environments. Although flexible and scalable, these databases often require users to understand their structure and specific query language, which can be difficult for non-experts. In this article, we propose an intelligent interface that enables users to interact with NoSQL databases. Our solution relies on query translation using advanced linguistic operations. This approach ensures domain independence, meaning that the interface does not depend on the specific vocabulary of the underlying database. Indeed, in many large-scale databases, users do not always know the exact names of collections (tables) and attributes. Even if a query is syntactically correct, it may return no results if the user employs terms different from those defined in the database. Our solution overcomes this limitation by analyzing the semantic context of queries and dynamically mapping them to NoSQL database schemas. This allows users to express their queries more naturally without needing to know the precise internal structure of the database. Moreover, our approach integrates an adaptive machine learning mechanism that incrementally improves the knowledge base with each user interaction, allowing the system to generalize to new queries and domains. This adaptive capability reduces the need for frequent reconfiguration and ensures optimal scalability The Experimental results show that our interface significantly reduces query formulation time and enhances the accessibility of NoSQL databases for non-specialists. By combining linguistic operations, domain independence, and adaptive learning, our solution facilitates NoSQL database utilization and optimizes the user experience.