Conversational Dashboard Generation with Explainability Analysis
摘要
The study shows new business intelligence(BI) system is designed by keeping in mind small-to-medium sized enterprises(SMBs). Methods such as Named Entity Recognition model based on DistilBERT is used to automatically parse user queries by extracting key entities the system will generate insightful visualization dynamically, enabling seamless dashboard creation. An integrated explainability module run by RAG provides detailed insights for visualization. For Role-Based Access Control (RBAC) JWT authentication is implemented with MySQL data store to ensure privacy and smooth management. This approach is consistent with growing BI trends, transforming raw data into business assets.