This paper proposes a philosophical reconstruction of the concept of “research data” in the context of Open Science. While institutional definitions vary widely and lack conceptual clarity, this study draws on insights from the philosophy of science and information to analyze three major perspectives—representational, relational, and diaphoric. Based on the latter, it introduces a compositional model that layers optional attributes (e.g., factive, retrievable, machine-readable) upon minimal differences. This model visualizes the structural assumptions behind institutional policies and serves as a bridge between conceptual analysis and data governance. It also clarifies how philosophical tools can inform and improve research data policies.
啄杜 榎本 (2026) studied this question.