Disentangling the schema turn: Restoring the information base to conceptual modelling

2025-09-01

AI Summary

This paper critiques the dominant “pattern-oriented paradigm” in conceptual modeling—characterized by an overemphasis on abstract conceptual patterns detached from their underlying information substrates, thereby undermining empirical grounding. To address this, we propose a novel paradigm that integrates conceptual patterns with their information substrates, supported by a pipeline-based technical framework enabling data-driven, automated, and verifiable modeling. Our contributions are threefold: (1) we shift away from traditional pattern-centricity by recentering the information substrate as a core modeling concern; (2) we demonstrate how contemporary technologies—including semantic parsing, ontology reasoning, and data-aware tooling—can concretely realize this paradigm; and (3) we validate its feasibility and effectiveness through the bCLEARer prototype system. The work extends both the theoretical foundations and practical scope of conceptual modeling, offering new avenues for database design, knowledge modeling, and model-driven engineering.

Technology Category

• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning
• Data Mining & Knowledge Management: Rule Mining & Pattern Mining
• Knowledge Representation and Reasoning: Ontologies

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Practical large-scale studies of user experience
• Web Mining and Content Analysis: Models for Web evolution

Abstract

If one looks at contemporary mainstream development practices for conceptual modelling in computer science, these so clearly focus on a conceptual schema completely separated from its information base that the conceptual schema is often just called the conceptual model. These schema-centric practices are crystallized in almost every database textbook. We call this strong, almost universal, bias towards conceptual schemas the schema turn. The focus of this paper is on disentangling this turn within (computer science) conceptual modeling. It aims to shed some light on how it emerged and so show that it is not fundamental. To show that modern technology enables the adoption of an inclusive schema-and-base conceptual modelling approach, which in turn enables more automated, and empirically motivated practices. And to show, more generally, the space of possible conceptual modelling practices is wider than currently assumed. It also uses the example of bCLEARer to show that the implementations in this wider space will probably need to rely on new pipeline-based conceptual modelling techniques. So, it is possible that the schema turn's complete exclusion of the information base could be merely a temporary evolutionary detour.

Problem

• Critiquing schema-centric bias in conceptual modeling practices
• Exploring historical emergence and non-fundamental nature of schema turn
• Proposing inclusive schema-and-base approach with modern technology

Innovation

• Inclusive schema-and-base conceptual modelling approach
• Pipeline-based conceptual modelling techniques
• Automated empirically motivated practices