Article Per Year (5 Year)
p-Index From 2021 - 2026
0.562
P-Index
This Author published in this journals
All Journal Formosa Journal of Multidisciplinary Research (FJMR)
Iffat Sania Hossain
California State University
Civil Engineering, Building, Construction & Architecture Computer Science & IT Decision Sciences, Operations Research & Management Education Social Sciences

Published : 3 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 3 Documents
Search

 1 
AI-Powered Framework for Real-time Threat Detection and Response in Cloud Infrastructure Zakir Hossain; Md Emran Hossain; Nisher Ahmed; Md Farhad Kabir; Iffat Sania Hossain
Formosa Journal of Multidisciplinary Research Vol. 4 No. 4 (2025): April 2025
Publisher : PT FORMOSA CENDEKIA GLOBAL

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55927/fjmr.v4i4.168

Abstract

As most organizations worldwide  embrace cloud computing services, cloud infrastructure security has become a significant concern. With cybersecurity attacks changing at an unprecedented rate in the  cloud environment, the methods for detection and response must become more robust. This  study presents an AI based framework to enhance the real-time detection and response to threats in cloud infrastructure. A possible threat that, if in a real-world scenario, could and would have been detected in real-time and was detected using clustering on e huge amount of cloud traffic. AI algorithms that detect malicious behaviour also assist in calculating the severity of the threat and recommend some flip of  a switch to change things instantly.  At the heart of the framework is its capacity for cumulative learning about new data, adjusting to emerging attack patterns and achieving low false positive rates. Additionally, it uses a hybrid approach that combines signature based detection  with anomaly detection to prevent known and unknown threats. Using this combination, the framework can detect new attack vectors that may be overlooked by traditional means.
Assessing the Potential and Ethical Implications of Agentic AI in Surveillance Technology Nisher Ahmed; Md Emran Hossain; Zakir Hossain; Md Farhad Kabir; Iffat Sania Hossain
Formosa Journal of Multidisciplinary Research Vol. 4 No. 4 (2025): April 2025
Publisher : PT FORMOSA CENDEKIA GLOBAL

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55927/fjmr.v4i4.167

Abstract

Agentic AI creates a system rather than a tool, an autonomous entity interacting with the world as humans do. By functionalizing agentic AI into surveillance technologies, we can increase surveillance systems' efficiency, accuracy, and scope, enabling them to monitor vast expanses of public space or issue, for example, a summary of dialogue from thousands of social media posts. However, it also interrogates the ethical dimensions of these systems, including the possible loss of privacy, accountability, and bias in decision making. AI surveillance technology is in use everywhere, but creating a surveillance state at the expense of civil rights, privacy, and freedom is a problem that can't be solved, no matter how many people are willing to use it. This paper proposes that despite the significant security operations improvements realized through agentic AI, the technology warrants an ethical framework with appropriate regulatory guardrails to manage the accompanying risks. Some points raised include the importance of transparency, fairness and accountability in using AI in surveillance settings, ensuring that these technologies are employed in a responsible and human rights compliant manner.
Analyzing and Predicting Emotional Responses in Cyber Bullying Cases: A Deep Learning Approach Nisher Ahmed; Md Emran Hossain; Zakir Hossain; Md Farhad Kabir; Iffat Sania Hossain
Formosa Journal of Multidisciplinary Research Vol. 4 No. 4 (2025): April 2025
Publisher : PT FORMOSA CENDEKIA GLOBAL

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55927/fjmr.v4i4.166

Abstract

Cyberbullying is a threat on any  digital platform, and it can have a very harmful and emotional effect on the person receiving these types of comments. Here, we present a deep learning framework that utilizes NLP-based and neural network-based approaches to analyze and predict the  emotional responses associated with cyberbullying incidents. The model is trained and evaluated using a curated dataset of social media posts labeled  with emotions like anger, sadness, fear, and neutrality. Tokenization, lemmatization, and word embeddings (GloVe, BERT, etc.) are  the different preprocessing methods used to represent textual data. Write. Multiple architectures, such as CNNs, LSTM networks,  and transformer-based approaches, are compared to achieve high accuracy in emotional response classification. Experimental results  show that transformer models outperform traditional learning models for precision and recall. The results can lead to intelligent monitoring systems that identify harmful emotional content  followed by necessary, timely interventions. Such  research shows promise for AI-driven emotion analysis to support safer online environments.
Co-Authors Md Emran Hossain Md Farhad Kabir Nisher Ahmed Zakir Hossain