Behnam Mohammadi
This paper explores a model of human decision-making when faced with limited resources, such as time, illustrated by the scenario of browsing through numerous news articles on Google News. The study reveals a tradeoff between the potential gains from selecting "better" options and the effort involved in identifying them, a concept known as bounded rationality. This dynamic results in rational yet suboptimal choices, wherein readers may overlook news articles that could have provided the most satisfaction. By formulating this model as an optimization problem, a choice probability with Gibbs distribution emerges as its solution. Notably, traditional structural logit models can be viewed as a specific instance of this broader framework. The paper introduces a methodology for inferring individual consumer preferences in both static and dynamic contexts, demonstrating that an optimal consumer policy can be established even in the face of suboptimal decision-making. The findings presented hold practical implications for enhancing the effectiveness of personalized recommender systems, such as news aggregation platforms, by acknowledging and accommodating human fallibility in decision-making processes.