The societal system is an intricate composition of individuals, each contributing their distinct demographics and experiences. It is not just a collection of people; it is an interconnected network that goes beyond the sum of its separate parts. Within this network, even a marginal action can have ripple effects, leading to the diffusion of behaviors, information, or, as painfully evidenced, infectious diseases. These intricate connections introduce significant challenges to the realm of operations management, including increased decision-making complexity, data overload in analysis, and a lack of theoretical guidance. All these challenges come together to form the central question that runs through my research: "How can we leverage the vast wealth of data and information available to navigate this intricate societal system for more effective operational decision-making?"

In response to this growing need, my dissertation contributes to the intersection of social network analytics and operations management. The objective is to create a more precise reflection of our interconnected societal systems, which, in turn, enables improved decision-making across a broad spectrum of platforms. To this end, I have employed a diverse tool set. These include optimization for high-quality problem-solving, data analytics to uncover actionable insights, machine learning to enable data-driven decision-making, network and graph theory to better understand the interconnected systems, and stochastic simulation for informed evaluation, etc.

The dissertation comprises three papers that each examine a different facet of integrating social network analytics with operations management. In Chapter 2, we explore the promotion optimization strategy with the consideration of the diffusion effects, drawing on extensive data from a large-scale online platform. In Chapter 3, we propose a general approximation framework to evaluate the impact of nonprogressive diffusion, delving into both its theoretical underpinnings and practical applications. In Chapter 4, we highlight the significant findings and set the stage for future research, particularly focusing on the challenges of learning user behavior within a social network with limited data availability.