KSK. Swetha

K. Swetha

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


K. Swetha has a total of 1 co-authors. Here are the top co-authors by number of shared publications:

  • K. Sivaraman

    Department of Computer Science and Engineering, Bharath Institute of Higher Education and Research (BIHER), Chennai, Tamilnadu, India

    1 shared publication

Latest Publications

  • Open Access

    Advances in Nonlinear Variational Analysis for Cyber Criminal Detection via Smartphone Communication Patterns

    by K. Swetha

    The increasing dependence on digital communication needs sophisticated ways for identifying cybercrime, particularly on mobile devices. This paper provides a comprehensive methodology that uses Natural Language Processing (NLP) tools to analyze communication patterns and detect suspicious activity. To generate communication patterns, text is preprocessed, vectorized with TF-IDF, metadata is extracted, and clustering is performed. Anomalies are identified by calculating deviations from these patterns, and suspicious signals are reported for further study. The proposed strategy outperformed baseline techniques in terms of accuracy, precision, recall, and the F1 score. Statistical study supported the method's performance, highlighting its potential as a significant tool for improving cybersecurity measures. In addition, the methodology employs Nonlinear Variational Inequalities to improve anomaly detection in communication patterns, addressing deviations more robustly within a cybersecurity framework

  • Open Access

    SMART Model: A Robust Approach for Cyber Criminal Identification using Smartphone Data

    by K. Swetha, K. Sivaraman

    The SMART (Smartphone Metadata Analysis for Recognizing Threats) model is a novel approach to the identification of prospective cyber criminals by analyzing smartphone data, with a particular emphasis on social media interactions, messages, and call logs. The SMART model, in contrast to conventional methods that depend on a wide variety of features, prioritizes critical parameters to ensure more precise and effective analysis. This model exhibits exceptional adaptability and robustness in a variety of data environments by employing sophisticated feature extraction and classification algorithms. This targeted approach not only improves the precision of threat identification but also offers a practicable solution for real-world cybersecurity applications, where data quality and consistency may vary.