Type: article
Abstract: Decalens: A Perspective-Method Separation Approach to AI Analysis Framework Generation A Conceptual Design Study Based on the Perspective Theory of Dao De Jing Abstract Purpose: To address the limitations of traditional prompt engineering—poor template reusability and constrained user methods—this paper proposes Decalens, an AI analysis framework generation methodology derived from the philosophy of Dao De Jing. Methods: The 81 chapters of Dao De Jing are grouped into 20 sets based on causal progression, then consolidated into 10 structured analysis perspectives via same-layer merging. A perspective-method separation mechanism is designed, where perspectives are fixed while methods are openly filled by users according to their own needs, generating personalized analysis frameworks, models, or engines. Results: Decalens supports 1,023 non-empty perspective combinations, each of which can be filled with unlimited methods. A five-layer interlocking structural consistency mechanism is designed to ensure logical coherence. Conclusions: Decalens transfers the generative authority of analysis frameworks from template designers to users, providing an extensible and composable methodology for non-technical users to conduct deep AI analysis. This paper is positioned as a conceptual design study and does not report empirical results; effectiveness awaits subsequent experimental validation. Keywords: Decalens; Dao De Jing; prompt engineering; perspective-method separation; analysis framework generation; conceptual design.
Author: Xu, Zhi Hao
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7213474976