The Interaction-based Caching (IC) theory, a novel approach developed in this research, enhances the retrieval of contents in ICN. It leverages the comprehensive interaction models of particle dynamics and quantum field theory to describe the complex interplay among caching parties in a paraphrased context. It organizes caching entities and their features within an extendible matrix, facilitating their differentiation, classification, and unification from multiple perspectives. The IC amalgamates the weighted relevant content-, client-, and node-related features, and entangles them with the network topology to optimize the spatiotemporal propagation of requests and replication of contents across the network. The resultant mechanisms enable the network to respond dynamically to clients’ demand patterns with an extensive range of nuanced prioritization policies and services. The IC’s framework is characterized by a flexible architecture that seamlessly integrates new entities, features, and functionalities. It consists of a logical formalism that describes the interactions among caching participants, remaining steadfast under variations of underlying axioms and assumptions, and functional algorithms that convert strategic principles into tactical actions tailored to specific contexts. The IC dynamically, deterministically, and organically optimizes content accessibility within an ICN network, consistently and coherently, considering the finite storage capacities and the interplay between various entities. It can encapsulate and reconcile diverse caching concepts within its panoramic scope. IC’s performance, evaluated in simulations in various scenarios, outperforms ICN’s standard strategy, LCE, regarding cache hit ratio, path length stretch, and content redundancy.