Improve the Bi-LSTM Model of University Financial Information Management Platform Construction

Abstract

Information Management Systems are vital tools for businesses and organizations to manage their data effectively, make informed decisions, ensure data security and compliance, and enhance operational efficiency. This paper presents a robust approach to information management through the utilization of the Optimized Probabilistic Bidirectional Long Short-Term Memory (OP_Bi-LSTM) model. The study developed into the architecture and application of Simulated Annealing (SA) for hyperparameter optimization, emphasizing the impact of fine-tuning on model performance. OP_Bi-LSTM, rooted in Bidirectional Long Short-Term Memory, exhibits superior sequential data processing capabilities, making it well-suited for a variety of information management tasks. The proposed OP_BI-LSTM hierarchical architecture further enhances its pattern recognition capabilities in the information management system. Results from extensive experimentation demonstrate the model's versatility and adaptability in various applications, such as sentiment analysis, fraud detection, and time-series forecasting. The proposed OP_BI-LSTM model performance analysis is evaluated for the consideration of the different customers and product data. The propsoed OP_Bi-LSTM model achieves the classification accuracy of 0.98. It is stated that OP_Bi-LSTM emerges as a powerful tool for information management, offering exceptional performance and adaptability. As the field of deep learning and information management continues to evolve, this model holds great promise for addressing complex data challenges and facilitating data-driven decision-making in a variety of industries.