Towards the Right Direction in BiDirectional User Interfaces

Title of the Paper

Towards the Right Direction in BiDirectional User Interfaces

Paper Information

• Research Area: Human-Computer Interaction (HCI), Bidirectional User Interface Design
• Keywords: Bidirectional design, BiDi, Right-to-left design, Bidirectional interface, Localization, UI element direction

Research Background and Problem

• Problem and Challenges: Globally, there are over 500 million users of bidirectional languages (e.g., Arabic, Hebrew, Urdu). These languages combine native right-to-left (RTL) text with left-to-right (LTR) Latin letters, numbers, and other characters, leading to directional inconsistencies in user interfaces (UIs) that can cause user confusion and errors. While major software providers have issued design guidelines addressing this issue, these guidelines are not comprehensive, and many UI elements still exhibit directional inconsistencies and errors.
• Significance: Inconsistent UI directionality affects users' automatic perception of information, increases learning and interaction time, induces ambiguity and misinterpretation, and reduces user satisfaction with the interface.
• Research Motivation and Related Work: Empirical studies on bidirectional user interfaces are relatively scarce. Although existing guidelines and standards (e.g., the Unicode Bidirectional Algorithm) provide basic references, researchers have found these guidelines to be insufficient and not fully reflective of the actual preferences and characteristics of bidirectional language users.

Proposed Solution

• Proposed Approach/Solution:
  1. Review existing bidirectional user interface design guidelines to analyze their shortcomings.
  2. Conduct field studies using Hebrew-based bidirectional user interfaces to identify common inconsistent UI elements.
  3. Investigate bidirectional language users' (particularly general users and HCI professionals) preferences for UI element directionality through user surveys and compare the results with existing design guidelines.
• Innovative Contributions: A methodological framework combining theoretical analysis and empirical research is proposed to systematically identify challenges in bidirectional design and provide new insights into interface directionality based on actual user preferences.
• Implementation Steps and Techniques:
  1. Review the top 200 Hebrew-language websites to identify common errors and directional inconsistencies in UI elements.
  2. Design a user survey featuring 27 UI elements with potential directional controversies to understand user preferences for different directions.
  3. Analyze the data using unsupervised clustering methods (K-Means), logistic regression, and statistical analyses (e.g., Z-tests, Mann-Whitney tests).

Research Findings

• Key Findings:
  1. Deficiencies in Design Guidelines: The study revealed that current bidirectional language design standards lack specific directional recommendations for 20 out of 32 UI elements.
  2. Diverse User Preferences: Among 27 UI elements, users clearly preferred the LTR version for 13 elements, the RTL version for 8 elements, and showed no significant preference for 6 elements.
  3. Differences Between HCI Professionals and General Users: HCI professionals generally favored adhering to RTL design standards, but their preferences differed from those of general users. For example, general users preferred LTR-oriented web scrollbars, while professionals favored RTL.
  4. Information Classification and Clustering Analysis: By categorizing UI elements (e.g., numbers and sequences, temporal flows), the experiment demonstrated that preferences for UI elements could be grouped based on information type.
• Advantages Compared to Existing Work:
  • Provides user data-driven recommendations for improving existing guidelines, challenging the traditional approach of "simply mirroring LTR interfaces to RTL."
  • Establishes a new benchmark for classifying UI element information using categorization and clustering methods.
• Experimental and Evaluation Results:
  • Overall, users showed a stronger preference for LTR versions, possibly influenced by LTR language habits (e.g., numerical writing conventions, cross-cultural norms).
  • Directional preferences for certain elements (e.g., star rating systems, promotional tags) showed significant divergence.
• Limitations and Future Directions:
  • Limitations:
    1. The data sample primarily consisted of Hebrew-speaking users, lacking validation for other bidirectional languages (e.g., Arabic).
    2. Insufficient classification of background factors (e.g., professional experience, usage contexts); future studies should expand the scope of the survey.
  • Future Directions:
    1. Conduct cross-linguistic studies involving users of various bidirectional languages.
    2. Introduce methods such as eye-tracking to explore the cognitive load and information processing of UI elements in greater depth.
    3. Investigate the impact of context and background on directional preferences to provide more customized guidelines for UI design in different use cases.

Conclusion: Through user research and theoretical analysis, this paper highlights the complexities of bidirectional language UI design and proposes specific improvement recommendations. Its significant contribution to the HCI community lies in providing an empirical foundation for bidirectional interface design and advocating for broader research and practical support in this area.