Enhancing Enterprise Value Creation Through Intelligent Digital Transformation of the Value Chain: A Deep Learning and Edge Computing Approach

Ruiqing Liu,Yonghong Wang

DOI: https://doi.org/10.1007/s13132-024-02087-5

IF: 1.815

2024-05-25

Journal of the Knowledge Economy

Abstract:In the rapidly evolving landscape of sustainable science and technology development, the integration of intelligent and automated methods has become instrumental in enhancing production efficiency across various industries. A key driver of success in this new economic era is the innovative creation and application of business models by emerging enterprises, leading to substantial profit returns. Business models are fundamentally rooted in the logic of enterprise value creation, often examined through the lens of the value chain. This paper explores the intersection of deep learning, edge computing, and the value chain to optimize enterprise business models in the era of artificial intelligence. Drawing from the foundation of Michael E. Porter's value chain theory, this research delves into the integration of big data with the value chain to propose novel digital operation models for business optimization. However, it highlights the limitations of such approaches and underscores the need for real-time reflection and faster computing speeds, particularly in the context of AI-driven transformations. To address these challenges, we present a task processing framework that leverages the deep learning model Xception and Edge Computing (EC) technology. An innovative optimization algorithm rooted in EC technology is introduced to enhance the efficiency and energy consumption of deep learning models. This algorithm utilizes an exit point strategy for on-demand model optimization, significantly improving real-time performance. Also, a value chain task processing framework, which combines deep learning and EC technology, is proposed to efficiently analyze and process computing tasks within the value chain. This framework optimally configures Xception models and edge computing nodes across various value chain modules, mitigating high computing costs and boosting real-time performance. The experimental results validate the effectiveness of the proposed approach in meeting real-time and stability requirements for value chain computing. This research holds significant implications for business model transformation and value chain digitalization, offering a pathway to enhance enterprise efficiency and reduce workforce demands. Future work will extend the optimization algorithm to other deep learning models and further tailor the task framework to different value chain types, enhancing accuracy and efficiency in larger-scale business operations.

economics