Chih-Yueh Chou

Students’ after-class self-evaluated comments are useful for understanding students’ learning and reflecting teacher’s teaching. Researchers and engineers have attempted to apply educational data mining techniques, such as text analysis, sentiment analysis, machine learning, and deep learning to develop classification mechanisms of students’ self-evaluated comments. This study was carried out to develop aspect and sentiment classification mechanisms to automatically classify students’ self-evaluated comments into seven aspect categories and three sentiment categories. We investigated the impact of whether we should exclude nonsense data or not, the impact of different feature extraction methods, and the impact of different classification models on classification accuracy. The results showed that the combination of bidirectional encoder representations from transformers (BERT) word embedding feature extraction and Random Forest classification showed the best accuracy (90.7%) on aspect classification when including nonsense data, whereas the combination of BERT-word embedding feature extraction and Random Forest classification had the best accuracy (93.2%) on aspect classification when excluding nonsense data. Including nonsense data reduced the classification accuracies. In addition, the combination of one-word bag-of-words feature extraction and Random Forest classification presented the best accuracy (99.5%) with regard to sentiment classification.

In self-regulated learning (SRL), students organize, monitor, direct, and regulate their learning. In SRL, monitoring plays a critical role in generating internal feedback and thus adopting appropriate regulations. However, students may have poor SRL processes and performance due to their poor monitoring. Researchers have suggested providing external feedback to facilitate better student SRL. However, SRL involves many meta-cognitive internal processes that are hidden and difficult to observe and measure. This study proposed a SRL model to illustrate the relationship among external SRL tools, internal SRL processes, internal feedback, and external feedback. Based on the model, this study designed a system with SRL tools and open leaner models (OLMs) to assist students in conducting SRL, including self-assessing their initial learning performance (i.e. perceived initial performance and monitoring of learning performance) after listening to a teacher’s lecture, being assessed by and receiving external feedback from the OLM (i.e. actual performance) in the system, setting target goals (i.e. desired performance) of follow-up learning, conducting follow-up learning (i.e. strategy implementation), and evaluating their follow-up learning performance (i.e. perceived outcome performance and strategy outcome monitoring). These SRL tools also externalize students’ internal SRL processes and feedback, including perceived initial, desired, and perceived outcome performances, for investigation. In addition, this study explores the impact of external feedback from the OLM on students’ internal SRL processes and feedback. An evaluation was conducted to record and analyze students’ SRL processes and performance, and a questionnaire was administered to ask students about their SRL processes. There are three main findings. First, the results showed that students often have poor internal SRL processes and poor internal feedback, including poor self-assessment, inappropriate target goals, a failure to conduct follow-up learning, and a failure to achieve their goals. Second, the results revealed that the SRL tools and external feedback from the OLM assisted most students in SRL, including monitoring their learning performance, goal-setting, strategy implementation and monitoring, and strategy outcome monitoring. Third, some students still required further support for SRL.

A competency-based curriculum involves courses for cultivating core competencies required for specific professions to enable students to take courses to cultivate their core competencies. This paper presents a curriculum level and competency-based learning analytics dashboard system with simulatable open learner models (OLMs) of core competencies for assisting students in setting goals for course performance. At first, the system provides students with their OLMs of core competencies based on their taken courses and grades. After that, students are asked to set their goals for course grades of all their courses during the current semester. The system displays the simulation of students’ future OLMs if they achieve their goals for course grades this semester. Students can re-set their goals and conduct the simulation of OLMs again until they satisfy the simulation results. The system also enables students to set detailed goals for attendance, assignment hand-in rate, midterm exam, and final exam in order to achieve their goals for course grade. A preliminary evaluation was conducted. Most students agreed that the simulatable OLMs assisted them in understanding the influence of their goals for course grades on core competencies and in setting goals for course grades. Most students also stated that setting detailed course goals prompted them to achieve the goal for course grade.

A competency‐based curriculum involves an outcome‐based approach for cultivating graduates’ core competencies required for specific professions. A curriculum committee defines graduates’ outcomes of core competencies, designs a curriculum to cultivate core competencies, evaluates graduates’ outcomes, and reflects on and regulates the curriculum. However, curriculum committees lack systematic evaluative information for reflection. Learning analytics, an emerging data‐driven analytics of educational data, could be applied to assist curriculum committees in reflection. This study proposes competency‐based learning analytics, including seven analytic tools, to analyze curricula and graduates’ academic records to assist curriculum committees in reflecting (1) objectives, competencies, and curriculum design of competency‐based curricula and (2) faculty teaching and student learning. The proposed learning analytics were conducted on 14 departments of a university. This study reports the curriculum committees’ eight practical reflections of curricula, faculty teaching, and student learning. This study illustrates potential applications and impact of competency‐based learning analytics on competency‐based curricula.

Programming involves complex skills and knowledge. Novice students encounter many difficulties and lack information for reflecting and improving their proficiencies in programming skills and knowledge. This paper presents a work-in-progress research to design a system with an open student model based on an assessment framework of iteration programming. The assessment framework includes strategies of counter-controlled loop and sentinel-controlled loop, implementation skills of while, for, and do-while, and program tracing abilities. The system presents open student models to students in order to assist students in reflecting their proficiencies in iteration programming.