i-LaTeX: Manipulating Transitional Representations between LaTeX Code and Generated Documents

Prototyping & User TestingComputational Methods in HCIUI/UX DesignersHCI Researchers

Document Title

i-LATEX: Manipulating Transitional Representations between LaTeX Code and Generated Documents

Document Information

  • Field of Study: Human-Computer Interaction, design and evaluation of scientific document editing tools
  • Keywords: LaTeX, document editor, transitional representations, user experience, interaction design, tables, mathematical formulas, images, visual editing

Research Background and Issues

  • Identified Problems or Challenges:
    • Editing documents using LaTeX is complex: code editors only provide general functionalities, while WYSIWYG (What You See Is What You Get) interfaces support only partial language features.
    • The textual encoding abstraction of LaTeX and the mapping between source code and output documents pose challenges for users.
    • Editing complex structures such as formulas and tables is particularly time-consuming and prone to errors.
    • Long compilation times and the duality of code and output (code-PDF duality) increase the cost of modifications and debugging.
  • Significance:
    • LaTeX is an essential tool for scientific and technical document preparation, but its high complexity reduces user productivity.
    • Particularly in academic contexts, these issues affect the efficiency of a large number of users.
  • Research Motivation and Related Work:
    • Through interviews with 11 users, the authors identified common pain points for LaTeX users, such as difficulty in locating code, a steep learning curve, and challenges in editing specific elements (e.g., tables, images).
    • Existing tools focus on either code or output but lack support for mapping between the two.
    • Inspired by document enhancement systems, hybrid programming environments, and output-oriented programming, the authors propose a more efficient interaction approach to address these issues.

Solution

  • Proposed Method or Solution:
    • Introduce "Transitional Representations," an interactive visual bridge between textual code and output documents.
    • Develop a LaTeX editor named i-LATEX, which supports transitional representations for mathematical formulas, tables, images, and grid layouts.
  • Innovative Aspects of the Solution:
    • Transitional representations offer a novel interaction model that provides consistent, bidirectional editing capabilities between code and output. Users can influence display content through code or adjust code via interactive modifications.
    • Supports more efficient localized real-time editing.
    • Four key features: locality (affects specific code segments only), bidirectionality (real-time synchronization between code and graphical representation), persistence (can be activated anytime), and embedding (directly binds elements in the output document).
  • Implementation Steps and Key Technologies:
    1. Use custom commands or environments in LaTeX documents, such as replacing the standard tabular environment with itabular to enable transitional representations for tables.
    2. Design an interactive UI directly linked to PDF output content, allowing users to click on an element to activate its interactive visual representation.
    3. Implement four types of transitional representations:
      • Mathematical formulas: View formula code snippets and associate them intuitively.
      • Tables: Support grid-based visualization, row/column insertion, deletion, and sorting.
      • Images: Enable resizing and cropping operations.
      • Grid layouts: Extend support for visualizing and manipulating grid structures.
    4. Extend an open-source editor (Visual Studio Code) to implement the tool, with added features like error prompts and prevention of invalid code changes.

Research Outcomes

  • Specific Achievements:
    1. Experimental Design and Results:
      • Conducted experimental evaluations with 16 participants.
      • Assessed task performance with and without transitional representations across 9 tasks.
      • For table and image tasks, transitional representations improved task completion speed (44% faster) and reduced compilation attempts (41% fewer).
      • For mathematical formulas, transitional representations enhanced editing efficiency for less experienced users but had minimal impact on experienced users.
    2. User Feedback:
      • Most participants reported reduced mental load, time pressure, and frustration due to transitional representations.
      • Over 80% of participants expressed a preference for using this tool in the future.
    3. User Strategy Improvements:
      • Users no longer relied on complex code strategies (e.g., external tools or manual adjustments for image cropping).
      • More direct interaction reduced trial-and-error cycles during code processing.
  • Advantages:
    • Simple operations with low learning costs.
    • Optimized for specific tasks, such as table row/column updates and image size adjustments.
    • Supports both guided learning for beginners and efficient editing for experts.
  • Limitations and Future Directions:
    • Limitations:
      • Currently does not support transitional representations for user-defined LaTeX commands.
      • Performance may degrade when editing large documents.
      • Some implementation issues exist, such as potential code-data desynchronization during rapid interactions.
    • Future Directions:
      • Extend support to other content types (e.g., TikZ drawings, chemical formulas).
      • Enhance existing representation functionalities (e.g., support for table cell merging).
      • Optimize interface design, allowing users to adjust the position and closing logic of transitional representations.
      • Promote adoption among a broader user base, such as integration with mainstream platforms like Overleaf.
      • Expand transitional representations to other document description languages (e.g., HTML/CSS) to support more application scenarios.

Quick Actions

Share

Share this page

ios_share

https://hci.top/en/papers/chi/71977/2022

AdRecommended

Learn AI Coding at CodeNow

open_in_newOpen DOI Link
DOI: https://dl.acm.org/doi/abs/10.1145/3491102.3517494
At a Glance

Paper Snapshot

fact_check
dataset
Source
CHI
calendar_month
Year
2022
emoji_events
Award
No award tagged
group
Authors
2 authors
sell
Subtopics
Prototyping & User Testing, Computational Methods in HCI
work
Professions
UI/UX Designers, HCI Researchers
article
Content Status
Full text indexed
hub
Related Papers
10 related papers