Mangolika Bhattacharya

Mangolika Bhattacharya

Auto-generated profile, based on publicly available information

Mangolika Bhattacharya

School of Information and Technology, Illinois State University, Normal, IL, USA

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
4 February 2025 in IEEE Access
Verified
Journal Article
Open access
MPMihai Penica
Mihai Penica
MBMangolika Bhattacharya
Mangolika Bhattacharya
WOWilliam O’bRien
William O’bRien
SMSean McGrath
Sean McGrath
MHMartin Hayes
Martin Hayes
EOEoin O'Connell
Eoin O'Connell
The convergence of the Internet of Things (IoT) and the construction industry represents a potent opportunity to fundamentally redefine the processes of constructing and maintaining buildings. The integration of IoT technology has the potential to revolutionize building safety, efficiency, and sustainability, offering a promising avenue for innovation and advancement within the industry. In this work, we propose an innovative design and implementation of an interoperable physical access control IoT-based system that can be easily integrated with existing access control security systems and used in various environments such as offices, homes, and public buildings. Based on our developed and implemented physical access control IoT-based system, we identified several challenges in the current state of access control systems. These challenges include difficulties in integrating different security systems, the absence of interoperable standards, and the need for further research to enhance the scalability and performance of such systems.

14 October 2026 in SSRN Electronic Journal
Verified
Preprint
Open access
AKAneesh Krishna
Aneesh Krishna
RBRudra Prasad Baksi
Rudra Prasad Baksi
MBMangolika Bhattacharya
Mangolika Bhattacharya
SUShambhu Upadhyaya
Shambhu Upadhyaya
Advanced Persistent Threat (APT) actors are stealthy, adaptive, and persistent, so they often evade traditional intrusion detection systems. The clearest evidence appears during the intrusion phase, so host events must be tracked and labeled individually. No public Windows dataset does so. We present Prowl, a Windows kernel-operation trace dataset of 4,411,161 labeled events. Of these, 79.4% are file and registry operations that can be mapped to native system calls while retaining their arguments. 18.4% are periodic process resource samples, and 2.2% are process, thread, and image lifecycle callbacks. A breach-and-attack-simulation platform ran adversary tradecraft in an instrumented sandbox while Process Monitor recorded every event, giving 20 attack captures and five baselines. Existing host datasets label per trace, sample, time window, or red-team report, so benign events inside an attack capture inherit malicious labels. We label each event by execution-context provenance. Process instances are rebuilt from process and parent identifiers and chronological order. An event is malicious when its process instance descends from the attack process tree. This preserves attribution under identifier reuse and captures activity through signed system binaries. The release includes structural validation, harness-artifact scrubbing, and a label-conflict audit that sets a 99.84% accuracy ceiling. Grouped evaluation reaches 0.998 ROC-AUC, and leave-one-capture-out 0.987 +/- 0.025. However, process identity drives these scores. Process-name lookup alone reaches 0.9988 ROC-AUC, and withholding test process names drops logistic regression to 0.895. We therefore recommend identity-blind held-out-process evaluation as the primary protocol for determining whether a model learns attack behavior rather than process identity.

1 January 2025
Verified
Report
Closed access
HXHaiyan Sally Xie
Haiyan Sally Xie
MBMangolika Bhattacharya
Mangolika Bhattacharya
The nomenclature table for the Mathematical Digital Twin project description systematically defines key variables, parameters, and mathematical symbols used throughout the research. The table is deposited for open access, allowing other researchers to reference and build upon the work to enhance transparency and reproducibility.

14 May 2026 in SSRN Electronic Journal
Verified
Preprint
Open access
AKAneesh Krishna
Aneesh Krishna
RBRudra Prasad Baksi
Rudra Prasad Baksi
MBMangolika Bhattacharya
Mangolika Bhattacharya
SUShambhu Upadhyaya
Shambhu Upadhyaya
Advanced Persistent Threat (APT) malware employs stealthy, adaptive, and long-duration attack strategies that challenge traditional intrusion detection systems (IDS). In this paper, we propose SCARLeT (System-Call Analysis for Robust AI Learning based detection of advanced persistent Threats), an intelligent host-based IDS that leverages machine learning (ML) and natural language processing (NLP) techniques to detect APT-type malware using system-call metadata. A large-scale dataset containing over 4.6 million system-call events is constructed using a commercial threat-emulation platform combined with diverse real-world APT-style malware campaigns.We develop and evaluate multiple IDS models including Support Vector Machine (SVM), Logistic Regression (LR), Gradient Boosted (GB) decision trees, Random Forest (RF), Naive Bayes Classifier (NBC), and a transformer-based BERT model under different feature representation strategies. Through systematic benchmarking, we show that dense embedding pipelines exhibit scalability limitations when applied to large APT datasets. In contrast, sparse TF--IDF representations combined with classical ML models achieve near-perfect detection performance, reaching accuracies up to 99.99% while maintaining substantially lower computational cost. Although the NLP-based BERT model achieves strong contextual performance (99.88% accuracy), it does not surpass the best-performing sparse lexical models in line-level classification settings.Our analysis highlights important trade-offs among accuracy, runtime efficiency, and deployability, demonstrating that carefully engineered classical ML approaches remain highly competitive for large-scale cybersecurity applications. The proposed SCARLeT framework provides practical guidance for designing scalable and efficient IDS solutions capable of defending against evolving APT threats.