ABSTRACT VIEW
Abstract NUM 1433

A WRITING PROCESS VISUALIZATION TOOL BASED ON FINE-GRAINED INPUT LOG ANALYSIS
E. Maekawa1, Y. Okano2, H. Murao2
1 Otemae University (JAPAN)
2 Kobe University (JAPAN)
In writing education, the final written product alone often fails to capture the cognitive processes, challenges, and revision strategies employed by learners during composition. To address this issue, we have developed an innovative tool that records and visualizes the writing process in detail. The tool captures timestamped logs of user interactions, including keystrokes, mouse movements, and memo entries, enabling analysis and playback of how a learner's text evolves over time. Unlike conventional approaches that focus solely on the final output, our system emphasizes the importance of the writing process itself.

Each keystroke is recorded along with its timestamp, the input character, and the resulting text state (e.g., currentText). Mouse interactions are also tracked to identify transitions between writing areas, such as moving between the main editor and a memo section. Additionally, user-written memos are timestamped and synchronized with other activities, offering insights into the writer’s cognitive and planning processes.

To support interpretation and reflection, the tool provides several visualization features. A timeline slider based on real time (rather than text position) allows smooth playback of the writing session, including periods of inactivity, thus making hesitation, pauses, and revision behaviors visible. Additional visualizations such as writing intensity heatmaps, phase-transition timelines, and cumulative keystroke graphs help highlight patterns in writing fluency, engagement, and editing.

Moreover, by visualizing the non-linear and often fragmented nature of composition, the tool encourages learners to recognize the educational value of their trial-and-error efforts. This perspective aligns with contemporary pedagogical values that prioritize process over product.

Future development will focus on automatic phase detection, AI-supported feedback generation, integration with learning management systems (LMS), multilingual writing support, and the exploration of applications for learners with diverse cognitive characteristics or learning needs.

This study presents a novel example of how detailed learning data and purposeful visualization design can enhance writing instruction. By bridging the gap between product and process, the tool supports deeper understanding, reflective learning, and educational growth in writing.

Keywords: Writing Process Visualization, Learning Analytics, Process-Oriented Feedback, Learning Support System, Metacognitive Reflection.

Event: ICERI2025
Track: Digital & Distance Learning
Session: Learning Analytics & Educational Data Mining
Session type: VIRTUAL