Leveraging big data analytics to optimize user interaction in power marketing management information platforms

Abstract

Power marketing management solutions are essential for optimizing energy distribution and consumption. These platforms are largely dependent on user interaction data, which can be complex and large. It is important to take advantage of major data analyses to increase the user’s engagement and increase energy efficiency. This research examines how big data analytics can be used to improve user interactions on powerful marketing platforms. The specific goal is to improve energy efficiency, personal services and dynamic marketing techniques based on user behavior analysis. The dataset contains consumer consumption patterns, pricing information, and interaction logs. Research uses big data approaches to evaluate massive volumes of user interaction data. Preprocessing entails cleaning data, normalizing it using Z-score normalization, and dealing with missing information to ensure high-quality input for analysis. K-means Clustering was used to categorize consumers based on the extracted features. This research suggested a novel Puma Optimizer Tuned Bidirectional Long Short-Term Memory Neural Network (PO-BiNet) technique for analyzing changes in user behavior over time and predicting power consumption patterns, which can help with grid scheduling and resource allocation. The proposed PO-BiNet method provides practical recommendations, such as top demand periods and identifying energy-saving opportunities. For Household 5 at a 60-min interval, PO-BiNet achieved superior performance with MAE of 0.036, RMSE of 0.049, MAPE of 4.73, and R² of 0.985, significantly outperforming existing methods. Dynamic pricing methods and individual energy-saving incentives were designed to increase the user’s engagement and reduce total energy consumption.