TL;DR
This paper proposes a "periodic table" for cybersecurity by breaking down complex concepts—like risks, controls, and properties—into a finite set of Atomic Security Concerns (ASCs). By treating these atoms as a foundational vocabulary, the author aims to move cybersecurity from a field reliant on expert intuition to one based on a formal, composable language.
What Problem Does This Paper Address?
Cybersecurity is currently a tower of Babel. Professionals in governance, risk management, and architecture use a fragmented mess of heterogeneous vocabularies and abstractions. While we intuitively understand how a specific control (like MFA) relates to a specific risk (like account takeover), these relationships are rarely formalized. Instead, they exist as "tacit knowledge"—expertise locked in the heads of practitioners.
The author argues that this lack of a common conceptual model makes it incredibly difficult to automate security reasoning, ensure consistency across different frameworks (like NIST vs. ISO), or create traceable links between high-level business requirements and low-level technical implementations.
Methods
The paper introduces a conceptual framework based on the Atomic Security Concern (ASC) hypothesis. Inspired by chemistry, it suggests that cybersecurity knowledge can be represented as compositions of fundamental elements.
To identify these "atoms," the author proposes three core principles:
- Decomposability: A concern is not atomic if it can be consistently represented as a combination of other concerns.
- Preservation of Security Semantics: A decomposition is only valid if the resulting parts collectively preserve the original cybersecurity meaning.
- Reusability Across Concepts: A true atomic concern should appear repeatedly across different domains (e.g., "Identity" appears in authentication, digital signatures, and non-repudiation).
Using these principles, the paper proposes an initial taxonomy of 11 candidate ASCs: Identity, Authenticity, Access, Privilege, Secret, Integrity, Trust, Traceability, Evidence, Availability, and Recovery.
Datasets
Not discussed in the paper. This is a conceptual and theoretical work focused on knowledge representation rather than an empirical study using external datasets.
Key Insights and Ideas
- The Chemistry Analogy: The most significant "aha moment" is the shift from viewing security concepts as standalone primitives to viewing them as "molecules." For example, "Authentication" isn't an atom; it's a molecule composed of Identity (who are you?) and Authenticity (can you prove it?).
- Semantic vs. Technical Atomicity: Atomicity here isn't about how simple a technology is, but whether its security meaning can be further divided. Encryption is a mechanism, but its underlying atomic concern is Secret.
- Cross-Domain Mapping: Because ASCs are independent of specific technologies or frameworks, they can bridge the gap between different silos. A "Security Operations Center (SOC)" team and a "Logging" control can finally be described using the same underlying vocabulary (Traceability and Evidence).
- Foundation for Reasoning: By representing concepts as sets of atoms, we can use set theory (shared concerns, inclusions, aggregations) to mathematically reason about security.
Key Results
The paper demonstrates the power of this approach by decomposing several standard cybersecurity concepts into their atomic parts.
Beyond the table above, the author shows how this mapping works across different categories:
- Security Properties: Confidentiality = Secret + Access.
- Security Risks: Unauthorized Data Disclosure = Threat against Secret + Access.
- Security Controls: Encryption = Protection of Secret.
This alignment explains why a specific control mitigates a specific risk: they share the same atomic concern.
Limitations and Future Work
The author is clear that this work is intentionally conceptual. The proposed taxonomy is illustrative, not exhaustive.
Key areas for future research include:
- Formalization: Developing the mathematical foundations and formal logic for composition rules.
- Validation: Testing the taxonomy against a much broader range of real-world architectures and regulatory frameworks to ensure completeness.
- Automation: Exploring how this foundational language could support AI-assisted cybersecurity decision-making and automated risk analysis.
Summary updated