Research on Entropy Reduction Mechanism Modeling and Optimization of Enterprise Architecture Systems Based on Data Mining

View at publisher Summarise with VIRAT
Authors: Jie Yang, Zhiqi Ao, Junyang Hu, Xiaolu Li, Chun Ke
Cited by: 0
References: 3

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Addressing the entropy increase problem in enterprise architecture (EA) during digital transformation caused by business expansion and module iteration (such as disordered business-data coupling and inefficient resource allocation), existing static management methods suffer from limitations such as lagging entropy source identification and singular optimization objectives. This paper proposes an EA- Entropy Reduce algorithm and entropy reduction mechanism based on data mining. First, an entropy quantification model integrating four dimensions—business, data, technology, and organization—is constructed to accurately characterize disorder. Second, a two-stage algorithm is designed: the entropy source dynamic identification module (EA-ERID) combines an improved Apriori algorithm with an LSTM time-series model to locate high-contribution entropy sources; the multi-objective optimization module (EA-EROP) generates a solution that balances entropy reduction effect, cost, and efficiency through an improved NSGA-II algorithm. An experiment based on 3 years of EA (Enterprise Architecture) operation data (1.2 million records) from 10 manufacturing enterprises showed that the total entropy reduction rate reached 42.3%, an improvement of 47.4% compared to static algorithms and 26.3% compared to single data mining algorithms; the time for entropy increase source identification was 12.5s, and the time for generating the optimization solution was 25.8s, representing reductions of 55.8% and 38.7% respectively; business response latency was reduced by 31.2%, and costs were controlled within budget by 15%. The research indicates that this algorithm can effectively solve the dynamic entropy increase problem in EA, providing technical support for digital management.

Fields of Study & Subjects

Information Technology Governance and Strategy Big Data and Business Intelligence Business Process Modeling and Analysis

Publication details

Journal / source
IEEE International Conference on Power, Electronics and Green Energy (ICPEGE)
Pages
1275–1280
Publication date
28 January 2026
Language
EN