Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{175558,
        author = {M.Sai Laxman Reddy and N.Sanjay and Rathod Sai Vamshi Krishna},
        title = {Deep Learning-based Security and Usability Analysis Framework for Mobile Android Applications},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {11},
        number = {11},
        pages = {3069-3074},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=175558},
        abstract = {Fast mobile applications raise the importance for building effective scalable evaluation methods so as to answer issues pertaining to security issues to even those regarding usability issues. Typically, most conventional evaluation approaches do not support depth scalability thus becoming important enough to handle through advanced techniques of machine learning. The article makes a new proposition where there are large language models that also use the LSTM network providing an overarching framework used when assessing Android applications. Specif- ically, we make use of LLMs such as GPT (Generative Pre-trained Trans- former) that supports security assessment tasks, which include vulner- ability detection, code review, and identifying the patterns of security- related behaviors in application behavior. Meanwhile, we employ LSTM networks trained on user interaction data to generate usability scores in reflection of the general experience of the user. This is what synergistic integration of LLMs for security evaluation and LSTMs for usability as- sessment can offer: a dual focus approach to provide a more holistic view of an application’s strengths and weaknesses. Our framework is designed to automate and scale the evaluation process, providing actionable in- sights to developers and stakeholders. In fact, our approach is validated by conducting extensive experiments on a diverse set of Android appli- cations, including security vulnerabilities and feedback from users. The results show that our approach is capable of giving accurate, timely, and cost-effective assessments of both security and usability, thus helping improve the overall evaluation process for mobile applications.},
        keywords = {Security Assessment · Usability Evaluation Deep Learning Models.},
        month = {April},
        }