An Automated Insight Generation and Multi-Dataset Comparative Analysis System for Business Intelligence Dashboards

Kamya Rajesh Karde
Student, Dr. D.Y. Patil Arts, Commerce and
Science College, Pimpri, Pune (411018)

Poonam Rajesh Sonawane
Student, Dr. D.Y. Patil Arts, Commerce and
Science College, Pimpri, Pune (411018)

Abstract— Business Intelligence (BI) dashboards are widely adopted for performance monitoring across organizations; however, they rely heavily on manual interpretation of visualized data. As data volumes increase, identifying hidden patterns, inefficiencies, and anomalies becomes complex and time consuming. This paper presents an automated insight generation and multi-dataset comparative analysis system for BI dashboards using digital marketing campaign data. The proposed system performs multi-dimensional analysis across channels, campaigns, demographics, and behavioral metrics to automatically generate explainable business insights. An anomaly detection mechanism using Isolation Forest identifies abnormal trends in key performance indicators. Insights are ranked based on business impact, and a recommendation engine provides actionable decision support. Additionally, the system supports comparative analysis across multiple datasets to assess robustness and sensitivity of generated insights. The framework aims to reduce manual analytical effort and enhance data-driven decision-making in enterprise environments.Keywords— Automated Insights, Business Intelligence, Augmented Analytics, Anomaly Detection, Data Analytics,Decision Support Systems