International Journal of Big Data Intelligent Technology, 2026, 7(3); doi: 10.38007/IJBDIT.2026.070303.
Zheng Li
Jacobs School of Engineering, University of California, San Diego, La Jolla, 92093, California, United States
Jetpack Compose; declarative user interface; Composition; runtime hotspots; optimized object selection
With the increasing application of Jetpack Compose in Android native UI development, state-driven updates reduce the complexity of UI construction. However, compile-time static risks and actual composition overhead are not always consistent. This paper proposes the Evidence-Guided Recomposition Optimization (EGRO) method, focusing on "Now in Android," separating hotspot discovery from performance evaluation, and comparing the original implementation, compiler single-evidence optimization, and EGRO. Results show that only 7 out of 31 static candidates become stable hotspots; EGRO reduces ECTD by 9.86%, yielding lower ECTD than the compiler single-evidence group, but the difference in P99 frame-level CPU generation time between G1 and G2 is not significant. The research indicates that runtime cost evidence can improve the targeting of optimization objects, while improvements within composition are not directly equivalent to frame-level performance improvements.
[1] Kwiatkowski A, Smołka J. Performance analysis of Jetpack Compose components in mobile applications[J]. Journal of Computer Sciences Institute, 2025, 37: 391–398. DOI: 10.35784/jcsi.7914.
[2] Lin H, Liu C, Li Z, et al. Aging or Glitching? What Leads to Poor Android Responsiveness and What Can We Do About It?[J]. IEEE Transactions on Mobile Computing, 2024, 23(2): 1521–1533. DOI: 10.1109/TMC.2023.3237716.
[3] Liao D, Pan S, Yang S, et al. A Comparative Study of Android Performance Issues in Real-World Applications and Literature[J]. ACM Transactions on Software Engineering and Methodology, 2026, 35(7): Article 212, 1–32. DOI: 10.1145/3771932.
[4] Lv M, Hu T, Cui M, et al. Improving UI responsiveness in Android by restructured rendering[J]. Journal of Systems Architecture, 2025, 168: Article 103580. DOI: 10.1016/j.sysarc.2025.103580.
[5] Rua R, Saraiva J. A large-scale empirical study on mobile performance: energy, run-time and memory[J]. Empirical Software Engineering, 2024, 29(1): Article 31. DOI: 10.1007/s10664-023-10391-y.
[6] Prestat D, Moha N, Villemaire R, Avellaneda F. DynAMICS: A Tool-Based Method for the Specification and Dynamic Detection of Android Behavioral Code Smells[J]. IEEE Transactions on Software Engineering, 2024, 50(4): 765–784. DOI: 10.1109/TSE.2024.3363223.
[7] Yu S, Fang C, Du M, et al. Practical, Automated Scenario-Based Mobile App Testing[J]. IEEE Transactions on Software Engineering, 2024, 50(7): 1949–1966. DOI: 10.1109/TSE.2024.3414672.
[8] Clerissi D, Denaro G, Mobilio M, Mariani L. DBInputs: Exploiting Persistent Data to Improve Automated GUI Testing[J]. IEEE Transactions on Software Engineering, 2024, 50(9): 2412–2436. DOI: 10.1109/TSE.2024.3439002.
[9] Callan J, Petke J. Multi-objective improvement of Android applications[J]. Automated Software Engineering, 2025, 32(1): Article 2. DOI: 10.1007/s10515-024-00472-7.
[10] Fawad M, Rasool G, Palma F. Android Source Code Smells: A Systematic Literature Review[J]. Software: Practice and Experience, 2025, 55(5): 847–882. DOI: 10.1002/spe.3401.
[11] Wang, Z. (2026). Construction of a Data-driven Decision-making Framework for Bulk Commodity Supply Chains. Engineering Advances, 6(3).
[12] Wang, Z. (2026). Research on Operation Data Analysis and Efficiency Improvement of the New Energy Industry Based on Machine Learning. Procedia Computer Science, 285, 605-612.