Abstract:
This study employs Fuzzy C-Means (FCM) clustering to analyze work-life boundary patterns, with direct applications in Library science and smart nation initiatives. By analyzing behavioural data, we identified three wellbeing profiles: socially fulfilled, high-stress/imbalanced, and balanced. Methodologically, we integrate dimensionality reduction techniques (PCA/ t-SNE) and multivariate visualization to affirm the validity of clusters. Crucially, we reframe findings through the perspective of library science: Clustering techniques to enable personalized user services in smart libraries by categorizing information-seeking behaviours; identified patterns can guide staff wellness programs to mitigate burnout in digital library environments and financial security indicators correlate with digital literacy adoption. Results reveal stress (r=0.78), social support (r=0.68), and financial security (r=0.72) as key differentiators. Moreover, the methodological innovations
include FCM parameter optimization (fuzziness m=2.1) and cluster validation by Fuzzy Partition Coefficient (FPC=0.709). This demonstrates how machine learning can transform library operations through data-driven user understanding and staff wellbeing management. The framework offers tools for libraries to segment users for tailored resource allocation, design workplace interventions, and promote national smart citizenry goals through behavioural informatics. While this study uses a general behavioural dataset as a proxy, in future study employing library-specific data to improve contextual relevance and operational applicability, thereby strengthening the research rigor.
