Erwin Lopez | Searcharxiv

arXiv2026

Is Log-Traced Engagement Enough? Extending Reading Analytics With Trait-Level Flow and Reading Strategy Metrics

Student engagement is a central construct in Learning Analytics, yet it is often operationalized through persistence indicators derived from logs, overlooking affective-cognitive states. Focusing on the analysis of reading logs, this study examines how trait-level flow - operationalized as the tendency to experience Deep Effortless Concentration (DEC) - and traces of reading strategies derived from e-book interaction data can extend traditional engagement indicators in explaining learning outcomes. We collected data from 100 students across two engineering courses, combining questionnaire measures of DEC with fine-grained reading logs. Correlation and regression analyses show that (1) DEC and traces of reading strategies explain substantial additional variance in grades beyond log-traced engagement ({\Delta}R2 = 21.3% over the baseline 25.5%), and (2) DEC moderates the relationship between reading behaviors and outcomes, indicating trait-sensitive differences in how log-derived indicators translate into performance. These findings suggest that, to support more equitable and personalized interventions, the analysis of reading logs should move beyond a one-size-fits-all interpretation and integrate personal traits with metrics that include behavioral and strategic measures of reading.

cs.CY↗

arXiv2024

Towards Semi-Autonomous Robotic Arm Manipulation Operator Intention Detection from Force Data

In hazardous environments like nuclear facilities, robotic systems are essential for executing tasks that would otherwise expose humans to dangerous radiation levels, which pose severe health risks and can be fatal. However, many operations in the nuclear environment require teleoperating robots, resulting in a significant cognitive load on operators as well as physical strain over extended periods of time. To address this challenge, we propose enhancing the teleoperation system with an assistive model capable of predicting operator intentions and dynamically adapting to their needs. The machine learning model processes robotic arm force data, analyzing spatiotemporal patterns to accurately detect the ongoing task before its completion. To support this approach, we collected a diverse dataset from teleoperation experiments involving glovebox tasks in nuclear applications. This dataset encompasses heterogeneous spatiotemporal data captured from the teleoperation system. We employ a hybrid Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) model to learn and forecast operator intentions based on the spatiotemporal data. By accurately predicting these intentions, the robot can execute tasks more efficiently and effectively, requiring minimal input from the operator. Our experiments validated the model using the dataset, focusing on tasks such as radiation surveys and object grasping. The proposed approach demonstrated an F1-score of 89% for task classification and an F1-score of 86% classification forecasted operator intentions over a 5-second window. These results highlight the potential of our method to improve the safety, precision, and efficiency of robotic operations in hazardous environments, thereby significantly reducing human radiation exposure.

cs.RO↗