Sandeep Bhattacharjee has a total of 11 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Deep Defender: An Adaptive Edge Cloud Framework for Intelligent DDoS Detection, Deception, and Tracking

    by Sandeep Bhattacharjee

    Cloud computing environments are increasingly vulnerable to Distributed Denial-of-Service (DDoS) attacks and sophisticated cyber intrusions that compromise service availability, scalability, and operational security. In this research, we have developed a resilient cloud-based intrusion detection framework using a hybrid Convolutional Neural Network–Long Short-Term Memory (CNN-LSTM) deep-learning architecture for intelligent DDoS detection and cyberattack tracking. The proposed framework integrates spatial feature extraction and temporal traffic-learning mechanisms to effectively capture complex network-behavior patterns within heterogeneous cloud environments. We used the NSL-KDD dataset for experimental validation, which included preprocessing, feature normalization, correlation analysis, feature-importance evaluation, PCA visualization, and t-SNE-based nonlinear traffic analysis. The results reflected high detection performance (accuracy-99.14%, precision-99.06%, recall-99.14%, and F1 score-99.04%) besides low false-positive behavior and strong attack-prediction. The developed framework exhibited robust adaptive intrusion-learning capability, reliable multi-class attack classification, and scalable operational suitability for intelligent cloud-security monitoring and resilient cyber-threat management applications.

  • Open Access

    A fuzzy–blockchain framework for crowdsourced fund allocation (FB-CFA) to startup MSMEs

    by Sandeep Bhattacharjee

    Purpose: Access to finance remains a major constraint for startup micro, small and medium enterprises (MSMEs), particularly within decentralized crowdfunding environments where uncertainty, fraud risk and inefficient fund allocation persist. Emerging technologies such as fuzzy logic and blockchain offer potential solutions to improve transparency, risk evaluation and automated funding decisions. This paper aims to develop and validate a fuzzy – blockchain framework for crowdsourced fund allocation to startup MSMEs by integrating fuzzy inference with blockchain-enabled smart contracts to support transparent, secure, and risk-aware funding decisions. Design/methodology/approach: This study proposes a hybrid fuzzy–blockchain framework for crowdsourced fund allocation to startup MSMEs. Financial inclusion indicators, including mobile money usage, digital payments, account ownership, credit access, crowdfunding activity, startup age and employee size, were collected from the Financial Inclusion Data Story GitHub repository, derived from the World Bank Global Findex Database. This method uses several components, including data normalization, correlation analysis, triangular fuzzy membership functions, a 27-rule Mamdani inference system and the verification of blockchain transactions and smart contracts, to manage fund allocation. Findings: The findings suggest that the proposed strategy leads to greater transparency and risk-aware allocation of funds across the Seed, Series A and Series B investment stages. Blockchain authentication manages transaction reliability while fuzzy inference processes can reduce uncertainty in funding decisions and fraud evaluations. Research limitations/implications: The results indicate that integrating fuzzy decision models with blockchain verification significantly improves transparency, reduces fraud vulnerability and optimizes funding allocation to MSMEs across different investment stages. Practical implications: This study relies on simulated crowdfunding transactions and secondary financial inclusion data sets, which may not fully capture real-time investor behavior or regulatory variations across crowdfunding platforms. Social implications: The framework provides fintech platforms and policymakers a flexible model to help make decisions about safe crowdfunding, allowing for automatic risk evaluation and clear tracking of investments in startups. Originality/value: The results illustrate the effectiveness of integrating fuzzy decision models with blockchain-based technology to improve trust, automation and governance in decentralized crowdfunding systems to promote the growth of MSMEs.

  • Open Access

    Discovering disease risk trends by applying Latent Dirichlet Allocation sentiment analysis to social media health discussions

    by Sandeep Bhattacharjee

    Social media platforms generate large volumes of health-related content that can provide valuable insights into emerging disease concerns and public perceptions. This study explores disease-risk trends through topic modeling and sentiment analysis of social media discussions. We used Latent Dirichlet Allocation (LDA) to extract topics and VADER sentiments from 50,000 statements related to diseases. The model performance was evaluated in terms of coherence, topic diversity, and stability metrics. The comparison was made with NMF and BERTopic. Three main topics were identified: COVID-19 symptoms and diagnosis; chronic disease management; and disease risk factors. The LDA model obtained coherence and topic diversity scores of 0.402 and 0.77, respectively. The sentiment analysis shows that the discussions are overall neutral, with some sentiments related to negative concerns. The results indicate that the use of topic modeling along with sentiment analysis is a useful approach to extract major disease-related topics from social media data. Future studies using transformer-based models, multilingual data, and longitudinal studies may improve digital disease surveillance and enable proactive public health monitoring.

  • Open Access

    A Novel Soft Computing-Driven Entropy-Fuzzy Rule Optimization Framework for Early Adaptive Intrusion Detection

    by Sandeep Bhattacharjee

    This paper proposes the Entropy-Fuzzy Rule Optimized Synergistic Tuning Framework for Intrusion Detection System (E-FROST IDS), an entropy-driven, fuzzy-rule-optimization-enhanced intrusion detection system designed to address challenges related to model accuracy, flexibility, and uncertainty. The results suggest that E-FROST improves the effectiveness of anomaly detection, reduces false alarms, and enhances interpretability through entropy-based uncertainty modeling. The novelty lies in integrating fuzzy entropy with an efficient decision engine tailored for dynamic threat environments. The main contribution is a unified and adaptive IDS capable of identifying threats across a wide range of network conditions. Major implications include strengthened cyber-defense preparedness, greater transparency for security analysts, and a flexible framework for future intelligent intrusion detection system solutions.

  • Open Access

    Sustainable Technological Expectations Using a New MCDM Approach: Sensitivity–Alternatives–Ranking Analysis (SARA)

    by Sandeep Bhattacharjee

    Multi-Criteria Decision-Making (MCDM) approaches that are transparent and robust are crucial to meet the rising demand for sustainable decision-making in technology management. This paper presents a novel methodology, the Sensitivity-Alternatives-Ranking Analysis (SARA) method, that combines sensitivity testing, alternative selection, and ranking visualization to assess long-term technology expectations. The suggested approach is implemented in a case study incorporating alternative technologies, and its results are compared with the VlseKriterijumska Optimizacija I Kompromisno Resenje (VIKOR) method and validated through extensive sensitivity analysis. The results demonstrate consistent and stable ranks across varying criteria weights, confirming the SARA framework's robustness. In comparison to standard MCDM methods, the SARA method enhances both interpretability and reliability through integration of technology evaluations, normalized scores, and visualization-based insights. The study suggests a practical and conceptually sound technique for long-term decision-making that may assist policymakers, researchers, and practitioners in conducting complex technical evaluations.

  • Open Access

    A Novel Multi Criteria Decision Framework for Modeling Trust and Loyalty in Social Media Marketing Ecosystems

    by Sandeep Bhattacharjee

    This study develops a fuzzy logic framework to quantify customer trust and loyalty in social media marketing, addressing uncertainty in online perceptions beyond traditional engagement metrics. In this study, we used the Mamdani fuzzy inference with centroid defuzzification for modelling interactivity, perceived trust, and SMM intensity variables. Results indicate non-linear relationships between moderate interactivity and high trust with an optimal loyalty score (7.8). It also reflects how higher interactivity and consistent social media engagement improve trust and loyalty outcomes. The framework captures subjective perception ambiguity but requires adaptive neuro-fuzzy optimization for broader generalizability. It offers marketers an interpretable decision-support tool promoting ethical, trust-based consumer relationships, bridging behavioral marketing and computational intelligence.


  • Open Access

    Machine Learning–Driven Sentiment Analysis for Tourism Insights: The Case of North-East Indian States

    by Sandeep Bhattacharjee

    This study examines publicly available online reviews from Sikkim, Meghalaya, Arunachal Pradesh, Assam, Nagaland, Mizoram, and Manipur to evaluate the effectiveness of sentiment analysis for improving tourism decision-making and addressing the constraints of traditional surveys in capturing emotional and informal experiential insights. Open-access datasets from TripAdvisor, Reddit, and travel blogs (GitHub) were methodically tokenized, lemmatized, and stop-word-free to ensure clean and analysable textual input. NRC Emotion Lexicon determined emotional tone, VADER and TextBlob estimated sentiment polarity, and a state-wise sentiment index was generated. Emotion and experience themes were clustered by K-Means. The findings suggest a considerably favourable attitude in Sikkim and Meghalaya due to natural beauty and hospitality, whereas Assam and Nagaland have mixed sentiments due to infrastructure constraints. Joy, trust, and surprise are dominant feelings, while sadness and contempt are related to service issues. The study concluded that AI-driven sentiment analytics enhances destination branding, service improvement, and policy planning, providing a strong foundation for data-informed tourist growth.

  • A Fuzzy-Probabilistic Model for Interpretable Entrepreneurial Orientation Prediction

    by Sandeep Bhattacharjee

    Purpose : The present study is an effort to improve entrepreneurship research by shifting from traditional theory-based and socio-cultural explanations to a data-driven, interpretable decision-modeling framework. It demonstrated the complex behavioral patterns that influence entrepreneurship skills using simple and transparent analytical techniques. Methodology : In this research, a new framework known as the hybrid fuzzy cognitive system–entrepreneurial decision model (FCSM-EDM) has been used. This framework used fuzzy logic to create rules based on entrepreneurship traits such as risk appetite, innovativeness, leadership skills, resilience, and business knowledge. It also utilized Shapley Additive exPlanations (SHAP)-based interpretability, Sankey visual reasoning, and explainable AI for drawing clear insights for entrepreneur decision-making. Findings : Results from the working model revealed a fuzzy accuracy of 79.99%, an interpretability score of 0.88, a precision of 0.75, and a high interpretability-precision tradeoff score (IPTS) of 0.809. The model exhibited RUC-AUC of 68%, a bootstrapped mean accuracy of 0.66, and a tiny standard deviation of 0.1999, which shows that it was both stable and reliable. Practical Implications : The model presented a precise and comprehensible decision-support system that allows educators, politicians, incubators, and financial institutions to evaluate entrepreneurial potential. It improved the design of evidence-based interventions, the identification of talent, and entrepreneurship training suited to various behavioral profiles. Originality : This study introduced a novel integration of fuzzy decision modeling, SHAP-based interpretability, and Sankey visual analytics in the field of entrepreneurship research. It offered a scalable and interpretable alternative for entrepreneurial and corporate decision-making applications by outperforming Bayesian fuzzy systems, fuzzy-ANFIS hybrids, random forests, and SVM-RBF models.

  • Open Access

    A Layered Entropy Model for Transparent Uncertainty Quantification in Medical AI: Advancing Trustworthy Decision Support in Small-Data Clinical Settings

    by Sandeep Bhattacharjee, Sanjib Biswas

    Smaller data environments with expert systems are generally driven by the need for interpretable reasoning frameworks, such as fuzzy rule-based systems (FRBS), which cannot often quantify epistemic uncertainty during decision-making. This study proposes a novel Layered Entropy Model (LEM) comprising three semantic layers: Membership Function Entropy (MFE), Rule Activation Entropy (RAE), and System Output Entropy (SOE). Shannon entropy is applied at each layer to enable granular diagnostic transparency throughout the inference process. The approach was evaluated using both synthetic simulations and a real-world case study on the PIMA Indian Diabetes dataset. In the real data experiment, the system produced sharp, fully confident decisions with zero entropy at all layers, yielding an Epistemic Confidence Index (ECI) of 1.0. The proposed framework maintains full compatibility with conventional Type-1 FRBS design while introducing a computationally efficient and fully interpretable uncertainty quantification capability. The results demonstrate that LEM can serve as a powerful tool for validating expert knowledge, auditing system transparency, and deployment in high-stakes, small-data decision domains, such as healthcare, safety, and finance. The model contributes directly to the goals of explainable artificial intelligence (XAI) by embedding uncertainty traceability within the reasoning process itself.