Internet of Digital Reality (IoD) is a technological vision that promises to radically transform existing digital ecosystems in a way that enables users to access all the content and capabilities - whether physical or digital - relevant to a goal-driven purpose in a highly integrated single environment. In this paper, we focus on a specific challenge that we expect will be crucial in making advances in this field: namely, the challenge of developing an effective search method that is personalized, adaptive and associative. As a possible solution to this challenge, we propose a graph-tensor based information model that incorporates the history of search keywords and inferred associations between them across potentially multiple search dimensions. We provide a brief discussion on why we assume this model to have advantageous properties and provide a short use-case example to motivate further research.