An Ontology of Dark Patterns Knowledge: Foundations, Definitions, and a Pathway for Shared Knowledge-Building

AI Ethics, Fairness & AccountabilityDark Patterns RecognitionUI/UX DesignersAI/ML Researchers & EngineersPrivacy Policy MakersHCI Researchers

Title of the Paper

An Ontology of Dark Patterns Knowledge: Foundations, Definitions, and a Pathway for Shared Knowledge-Building

Paper Information

  • Subject Area: Manipulative design patterns (dark patterns) in user interface design and interaction
  • Keywords: Dark patterns, deceptive design, user privacy, regulation, ontology, human-computer interaction

Research Background and Problem

  • Issues or Challenges:

    • Manipulative and deceptive design practices (e.g., "dark patterns") are becoming increasingly prevalent in digital environments, affecting user interactions on social media, e-commerce platforms, and mobile devices.
    • There is a lack of unified definitions for dark patterns, with terminological discrepancies across academic research, regulations, and industry practices. This limits the ability of academic research to support legal enforcement and the development of design standards.
    • Research on dark patterns is highly fragmented, making it difficult to track their evolution and achieve cross-disciplinary integration.
    • As new types continue to emerge across different fields, redundant efforts are increasing, exacerbating the problem of academic silos.
  • Significance:

    • Dark pattern designs threaten user autonomy and informed decision-making, with profound implications for user privacy, business integrity, and social equity.
    • A unified language and conceptual framework for dark patterns can help standardize digital design, promote interdisciplinary collaboration, and support legal and regulatory measures.
  • Research Motivation and Related Work:

    • The authors, through multiple academic engagements with designers, legal scholars, and regulators, identified the importance of establishing a shared language and centralized knowledge structure.
    • Existing studies (e.g., Mathur et al.) have preliminarily unified definitions and classifications of dark patterns, but challenges remain in practical applications and cross-disciplinary collaboration.

Solution

  • Method or Solution:

    • Proposes a three-layer ontology framework that categorizes dark patterns into high-level strategies, mid-level perspectives, and low-level practices.
    • Integrates 10 academic and regulatory classifications and systematically analyzes them to generate a unified ontology comprising 64 specific dark patterns.
    • Develops standardized syntactic structures for defining dark patterns and iteratively refines them through community feedback.
  • Innovations:

    • The ontology combines existing classifications from academic, regulatory, and practical domains, using a unified hierarchical modeling approach to analyze dark patterns.
    • Introduces "mid-level patterns" for the first time to bridge the gap between abstract strategies and concrete implementations.
    • Provides an extensible framework adaptable to specific applications in different domains or technologies.
  • Implementation Steps and Key Techniques:

    • Collects dark pattern classification literature from 10 sources and conducts content analysis.
    • Traces the origins of patterns to identify when and under what conditions they were first proposed.
    • Uses visualization tools (Miro) to cluster patterns and create mid-level patterns.
    • Validates the ontology within the community, involving researchers, designers, legal experts, and regulatory bodies.

Research Outcomes

  • Specific Outcomes:

    • Unifies the concept of dark patterns, proposing an ontology comprising 5 high-level patterns, 24 mid-level patterns, and 35 low-level patterns.
    • Provides standardized syntax and detailed definitions to express these patterns, covering relationships between new and existing patterns.
    • Develops open digital resources to support the expansion and sharing of the ontology by the public and researchers.
  • Advantages Compared to Existing Solutions:

    • Resolves inconsistencies in terminology across existing academic and regulatory classifications, improving the efficiency of cross-disciplinary collaboration.
    • Offers more granular pattern hierarchies, facilitating detection and analysis of dark patterns by researchers and regulators.
    • Establishes a traceable record of pattern evolution, aiding in understanding how dark patterns emerge and develop across different fields.
  • Experimental or Evaluation Results:

    • Preliminary evaluations within the academic community received positive feedback, highlighting the effectiveness and applicability of the pattern definitions.
    • The ontology has gradually been applied in practical research, such as mapping dark patterns to specific legal cases and regulatory measures.
  • Limitations and Future Directions:

    • Limitations:
      • The hierarchical classification of certain patterns may be contentious due to domain-specific differences.
      • The current version primarily relies on data sources from Western contexts, with limited consideration of cultural diversity and internationalization.
    • Future Directions:
      • Further refine definitions to address specific needs of legal and technical experts.
      • Expand the ontology to support new domains (e.g., virtual reality, healthcare applications) and cross-cultural studies of dark patterns.
      • Develop automated detection tools integrated with the ontology to support technical practices and real-time review by designers.

Conclusion

This paper proposes a novel three-layer ontology of dark patterns, incorporating standardized definitions and syntactic structures, and synthesizing academic and regulatory resources to provide a shared knowledge framework for studying and regulating dark patterns. The research outcomes facilitate interdisciplinary collaboration, enhance regulatory effectiveness, and offer a structured pathway and extensible solutions for future exploration of dark patterns.

Quick Actions

Share

Share this page

ios_share

https://hci.top/en/papers/chi/147221/2024

AdRecommended

Learn AI Coding at CodeNow

open_in_newOpen DOI Link
DOI: https://doi.org/10.1145/3613904.3642436
At a Glance

Paper Snapshot

fact_check
dataset
Source
CHI
calendar_month
Year
2024
emoji_events
Award
No award tagged
group
Authors
4 authors
sell
Subtopics
AI Ethics, Fairness & Accountability, Dark Patterns Recognition
work
Professions
UI/UX Designers, AI/ML Researchers & Engineers, Privacy Policy Makers, HCI Researchers
article
Content Status
Full text indexed
hub
Related Papers
1 related papers