Publishing type：Research paper (international conference proceedings)  

Understanding students' coding behaviors is crucial for providing targeted support in programming education. Automatic analysis of coding behaviors using machines can address the limitations of manual monitoring. Previous studies focused on coding behavior representations without considering differences relative to a model answer. We propose embedding vectors that capture these differences, enabling the distinction between simple and complex code solutions. Evaluating these vectors by predicting assignment scores, we achieved over 15% higher accuracy compared to conventional methods. This approach has the potential to enhance teachers' understanding of students' coding behaviors and improve support in programming education.

Scopus

researchmap

Publisher：31st International Conference on Computers in Education, ICCE 2023 - Proceedings  

Understanding students' coding behaviors is crucial for providing targeted support in programming education. Automatic analysis of coding behaviors using machines can address the limitations of manual monitoring. Previous studies focused on coding behavior representations without considering differences relative to a model answer. We propose embedding vectors that capture these differences, enabling the distinction between simple and complex code solutions. Evaluating these vectors by predicting assignment scores, we achieved over 15% higher accuracy compared to conventional methods. This approach has the potential to enhance teachers' understanding of students' coding behaviors and improve support in programming education.

Scopus

Language：English   Publishing type：Research paper (scientific journal)   Publisher：Springer Science and Business Media LLC  

Abstract

In this paper, we propose a cross-language font style transfer system that can synthesize a new font by observing only a few samples from another language. Automatic font synthesis is a challenging task and has attracted much research interest. Most previous works addressed this problem by transferring the style of the given subset to the content of unseen ones. Nevertheless, they only focused on the font style transfer in the same language. In many cases, we need to learn font style from one language and then apply it to other languages. Existing methods make this difficult to accomplish because of the abstraction of style and language differences. To address this problem, we specifically designed the network into a multi-level attention form to capture both local and global features of the font style. To validate the generative ability of our model, we constructed an experimental font dataset of 847 fonts, each containing English and Chinese characters with the same style. Results show that our model generates 80.3% of users’ preferred images compared with state-of-the-art models.

DOI： 10.1007/s10489-022-04375-6

Web of Science

Scopus

researchmap

Other Link： https://link.springer.com/article/10.1007/s10489-022-04375-6/fulltext.html

Publishing type：Research paper (scientific journal)   Publisher：Journal of Educational Computing Research  

This study helps improve the early prediction of student performance by RNN-FitNets, which applies knowledge distillation (KD) to the time series direction of the recurrent neural network (RNN) model. The RNN-FitNets replaces the teacher model in KD with “an RNN model with a long-term time-series in which the features during the entire course are inputted” and the student model in KD with “an RNN model with a short-term time-series in which only the features during the early stages are inputted.” As a result, the RNN model in the early stage was trained to output the same results as the more accurate RNN model in the later stages. The experiment compared RNN-FitNets with a normal RNN model on a dataset of 296 university students in total. The results showed that RNN-FitNets can improve early prediction. Moreover, the SHAP value was employed to explain the contribution of the input features to the prediction results by RNN-FitNets. It was shown that RNN-FitNets can consider the future effects of the input features from the early stages of the course.

DOI： 10.1177/07356331221129765

Web of Science

Scopus

researchmap

Publishing type：Research paper (scientific journal)   Publisher：IEEE Transactions on Learning Technologies  

In this study, we propose an integrated system to support learners' reviews. In the proposed system, the review dashboard is used to recommend review contents that are adaptive to the individual learner's level of understanding and to present other information that is useful for review. The pages of the digital learning materials that are estimated to be insufficiently understood by each learner and the webpages related to those pages are recommended. As a method for estimating such pages, we consider extracting the pages related to the questions that were answered incorrectly. We examined the accuracy of matching each question with the pages of the learning materials. We also conducted an experiment to verify the usefulness of the system and its effect on learning using a review dashboard. In the experiment, the evaluation of the review dashboard indicated that at least half of the participants found it useful for most types of feedback. In addition, the rate of change in quiz scores was significantly higher in the group using the review dashboard, which indicates that using the review dashboard has the effect of improving learning.

DOI： 10.1109/TLT.2022.3225206

Web of Science

Scopus

researchmap

Language：English   Publishing type：Research paper (scientific journal)  

Language：English   Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (international conference proceedings)   Publisher：IFIP Advances in Information and Communication Technology  

Teaching and learning from one another is one of the most effective ways for learners to acquire proactive learning attitudes. In this study, we propose a new learning support system that encourages mutual teaching and learning by introducing a mechanism that guarantees sustainability. Learners submit articles called “learning-articles” that summarize their own learning and knowledge. The proposed system not only accumulates and publishes these articles but also has a mechanism to encourage the submission of necessary topics. The proposed system has been in operation since the academic year 2020, and it has collected learning-articles across our university’s nine academic disciplines from more than 300 learners. To investigate the effects of sharing learning-articles on education from the learners’ perspectives, a questionnaire was distributed among 25 students.

DOI： 10.1007/978-3-031-43393-1_44

Scopus

researchmap

Publishing type：Research paper (international conference proceedings)  

researchmap

Other Link： https://dblp.uni-trier.de/rec/conf/edm/2023

Web of Science

Publishing type：Research paper (international conference proceedings)   Publisher：IFIP Advances in Information and Communication Technology  

Smart agriculture has assumed increasing importance due to the growing age of farmers and a shortage of farm leaders. In response, it is crucial to provide more opportunities to learn about smart agriculture at agricultural colleges and high schools, where new farmers are trained. In agricultural education, a system is used for managing environmental information, such as temperature and humidity, obtained from sensors installed in the field. However, it is difficult to make effective use of this system due to the time required to detect changes in the field interfering with class time and the problem of oversight. In this study, we proposed a field environment digest system that will help learners by providing the summarized field sensing information, and support them in analyzing the data. In addition, to examine the potential for using field sensing information in agricultural education, we investigated the usefulness of the summarized sensor information and students’ usage of this information. In this paper, we outline the contents of the developed system and the results of the digest evaluation experiments.

DOI： 10.1007/978-3-031-43393-1_10

Scopus

researchmap

Publisher：Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)  

When students use e-learning systems such as learning management systems and e-book systems, the operation logs are stored and analyzed to understand student learning behaviors. For implementing some applications, such as dashboard systems and at-risk student detection, the operation logs are mainly transformed into features designed by researchers. Such hand-crafted features, like the number of operations, are easily interpretable. However, the power of the hand-craft features may be limited for the recent large-scale educational dataset. In machine learning research, data-driven features are demonstrated to be a better representation than hand-crafted features. However, there are few discussions in the educational data due to a need for many operation logs. In this study, we collect reading logs of an e-book system. We propose a representation learning method for the reading logs based on contrastive learning. Our proposed method transforms time-series reading logs into reading behavior feature vectors directly without hand-crafted features. In our experiments, we demonstrate that the power of our feature representation is better than a traditional count-based hand-crafted feature representation in the at-risk student detection task. In addition, we investigate the characteristics of the feature space learned by our proposed method.

DOI： 10.1007/978-3-031-36272-9_35

Web of Science

Scopus

Publisher：30th International Conference on Computers in Education Conference, ICCE 2022 - Proceedings  

The use of e-book reading systems such as Bookroll, and their ability to record readers' activities allows the design of predictive models capable of identifying at-risk students from their reading characteristics. Even though previous works have obtained promising results in this task, these results may not evidence the expected prediction performance in practical applications due to their selected assessment methods. Accordingly, in this paper, we assess this performance in two practical scenarios. The first is when we keep stored data from previous years of our course which can be used to train our model, and the second is when we only have data from a different course to use in this training process. In order to obtain a more accurate assessment, we collected 92, 574 samples of predictive performances from different models under the above-mentioned conditions. We also considered different feature representations along with variational latent representations, which can leverage our previous data to automatically design general hidden features. From our results, we understand that in the first condition we can expect a relatively good predictive performance, especially when using variational latent representations. However, in the second condition we found that even when using them, the predictive performances are very limited resulting in an impractical solution.

Scopus

Publishing type：Research paper (international conference proceedings)  

In recent years, several kinds of e-learning systems, such as e-book and Learning Management System (LMS) have been widely used in the field of education. When students access these systems, their activities on the systems will be continuously and automatically recorded and stored as learning logs. As the learning logs are stored in association with students and indicate students' learning activities, most studies have been “student-based” learning log analyses focused on students and each student's learning behavior. However, the “student-based” learning log analysis focuses on each student's learning behavior during the entire lesson (for example, studied well or didn't study enough) and cannot show what they learned. Therefore, if there is a need to investigate students' learning behavior regarding each topic of the lesson, such as which topic is learned well and which not in order to optimize the syllabus, we cannot conduct “student-based” learning log analysis directly. Instead of “student-based” learning log analyses, this study describes a method of “learning-topic-based” learning log analysis. We will show how to convert a learning log associated with students into a learning-topic-associated one and shape the logs into a two-dimensional matrix of learning topics and learning activities. Then we apply Non-negative Matrix Factorization (NMF) to the matrix in order to extract the learning patterns by activity. In addition, we make a three-dimensional matrix (tensor) of students, learning topics, and learning activities by subdividing the learning activities of each learning topic by students. We then apply Non-negative Tensor Factorization (NTF) to the tensor to extract detailed learning patterns. The methods proposed in this study will help teachers to have a comprehensively view of students' learning behaviors towards each learning topic easily even if the learning log is in a large-scale, so teachers can adjust syllabus according to the attracted learning behaviors, which is helpful to increase learning efficiency.

Scopus

researchmap

Publisher：30th International Conference on Computers in Education Conference, ICCE 2022 - Proceedings  

In recent years, several kinds of e-learning systems, such as e-book and Learning Management System (LMS) have been widely used in the field of education. When students access these systems, their activities on the systems will be continuously and automatically recorded and stored as learning logs. As the learning logs are stored in association with students and indicate students' learning activities, most studies have been “student-based” learning log analyses focused on students and each student's learning behavior. However, the “student-based” learning log analysis focuses on each student's learning behavior during the entire lesson (for example, studied well or didn't study enough) and cannot show what they learned. Therefore, if there is a need to investigate students' learning behavior regarding each topic of the lesson, such as which topic is learned well and which not in order to optimize the syllabus, we cannot conduct “student-based” learning log analysis directly. Instead of “student-based” learning log analyses, this study describes a method of “learning-topic-based” learning log analysis. We will show how to convert a learning log associated with students into a learning-topic-associated one and shape the logs into a two-dimensional matrix of learning topics and learning activities. Then we apply Non-negative Matrix Factorization (NMF) to the matrix in order to extract the learning patterns by activity. In addition, we make a three-dimensional matrix (tensor) of students, learning topics, and learning activities by subdividing the learning activities of each learning topic by students. We then apply Non-negative Tensor Factorization (NTF) to the tensor to extract detailed learning patterns. The methods proposed in this study will help teachers to have a comprehensively view of students' learning behaviors towards each learning topic easily even if the learning log is in a large-scale, so teachers can adjust syllabus according to the attracted learning behaviors, which is helpful to increase learning efficiency.

Scopus

Publishing type：Research paper (international conference proceedings)  

Since the demand for programmers is increasing, programming courses are being offered widely. In this context, students' motivation can be damaged by difficulties they encounter in their programming courses. Although teachers' support is necessary to prevent such an issue, it is impossible for teachers to directly monitor all students' programming activities at the same time and determine which students have troubles with programming. Therefore, several studies have been conducted to help teachers monitor students. However, these studies do not provide an understanding of the activities of students who do not run their code, which may lead researchers to miss students who are in trouble. In this paper, we propose an indicator for detecting students who need coding support by analyzing programming logs that are recorded even when the students do not run their code. This gives teachers deeper insight into the students' programming performance. Although further work remains, the validation of this indicator shows that it could detect those students who are in trouble.

Scopus

researchmap

Publisher：30th International Conference on Computers in Education Conference, ICCE 2022 - Proceedings  

Since the demand for programmers is increasing, programming courses are being offered widely. In this context, students' motivation can be damaged by difficulties they encounter in their programming courses. Although teachers' support is necessary to prevent such an issue, it is impossible for teachers to directly monitor all students' programming activities at the same time and determine which students have troubles with programming. Therefore, several studies have been conducted to help teachers monitor students. However, these studies do not provide an understanding of the activities of students who do not run their code, which may lead researchers to miss students who are in trouble. In this paper, we propose an indicator for detecting students who need coding support by analyzing programming logs that are recorded even when the students do not run their code. This gives teachers deeper insight into the students' programming performance. Although further work remains, the validation of this indicator shows that it could detect those students who are in trouble.

Scopus

Publishing type：Research paper (international conference proceedings)  

The use of e-book reading systems such as Bookroll, and their ability to record readers' activities allows the design of predictive models capable of identifying at-risk students from their reading characteristics. Even though previous works have obtained promising results in this task, these results may not evidence the expected prediction performance in practical applications due to their selected assessment methods. Accordingly, in this paper, we assess this performance in two practical scenarios. The first is when we keep stored data from previous years of our course which can be used to train our model, and the second is when we only have data from a different course to use in this training process. In order to obtain a more accurate assessment, we collected 92, 574 samples of predictive performances from different models under the above-mentioned conditions. We also considered different feature representations along with variational latent representations, which can leverage our previous data to automatically design general hidden features. From our results, we understand that in the first condition we can expect a relatively good predictive performance, especially when using variational latent representations. However, in the second condition we found that even when using them, the predictive performances are very limited resulting in an impractical solution.

Scopus

researchmap

Publishing type：Research paper (scientific journal)   Publisher：Sustainability (Switzerland)  

The COVID-19 pandemic has led to an increase in online classes, and programming classes are no exception. In such a learning environment, understanding every student’s programming process is mostly impractical for teachers, despite its significance in supporting students. Giving teachers feedback on programming processes is a typical approach to the problem. However, few studies have focused on visual representations of the evolution process of source-code contents; it remains unclear what visual representation would be effective to this end and how teachers value such feedback. We propose two feedback tools for teachers. These tools visualize the temporal evolution of source-code contents at different granularities. An experiment was conducted in which several university teachers performed a user evaluation of the tools, particularly with regard to their usefulness for reviewing past programming classes taught by another teacher. Questionnaire results showed that these tools are helpful for understanding programming processes. The tools were also found to be complementary, with different aspects being highly evaluated. We successfully presented concrete visual representations of programming processes as well as their relative strengths and weaknesses for reviewing classes; this contribution may serve as a basis for future real-time use of these tools in class.

DOI： 10.3390/su14138084

Web of Science

Scopus

researchmap

Publishing type：Research paper (scientific journal)   Publisher：MDPI AG  

While positive effects of imitating other learners have been reported, the recent increases in the number of online classes have seriously limited opportunities to learn how others are learning. Providing information about others’ learning activities through dashboards could be a solution, but few studies have targeted learning activities on e-textbook systems; it remains unclear what information representations would be useful and how they would affect learning. We developed a dashboard system that enables live sharing of students’ learning activities on e-textbooks. An experiment was conducted applying the dashboard in an online class to evaluate its impact. The results of questionnaires and quizzes were analyzed along with learning activities on the e-textbook system. From the questionnaire results, the most useful feedback types were identified. Regarding the impact on learning, the study found that a higher percentage of students who used the dashboard followed the progress of the class than those who did not. The study also found that students who used the dashboard were more likely to achieve higher quiz scores than those who did not. This study is the first to reveal what specific feedback is useful and to successfully investigate the impact of its use on learning.

DOI： 10.3390/su14126946

Web of Science

Scopus

researchmap

Publishing type：Research paper (scientific journal)   Publisher：Springer Science and Business Media LLC  

<title>Abstract</title>With the increasing use of digital learning materials in higher education, the accumulated operational log data provide a unique opportunity to analyzing student learning behaviors and their effects on student learning performance to understand how students learn with e-books. Among the students’ reading behaviors interacting with e-book systems, we find that jump-back is a frequent and informative behavior type. In this paper, we aim to understand the student’s intention for a jump-back using user learning log data on the e-book materials of a course in our university. We at first formally define the “jump-back” behaviors that can be detected from the click event stream of slide reading and then systematically study the behaviors from different perspectives on the e-book event stream data. Finally, by sampling 22 learning materials, we identify six reading activity patterns that can explain jump backs. Our analysis provides an approach to enriching the understanding of e-book learning behaviors and informs design implications for e-book systems.

DOI： 10.1186/s40561-021-00183-6

Web of Science

Scopus

researchmap

Other Link： https://link.springer.com/article/10.1186/s40561-021-00183-6/fulltext.html

Web of Science

Publisher：Proceedings - 2022 IEEE International Conference on Teaching, Assessment and Learning for Engineering, TALE 2022  

Students' performance prediction is essential for instructors to observe each student's learning behavior to discover which students have become at-risk. Early prediction helps instructors to intervene in time and provide academic support to these students. However, instructors should grasp essential behavior points to survey students' academic performance. In this study, we propose the Scaled-Dot Product Attention that can mine the relationship between the student's learning behaviors and performance to find the essential features that directly affect students' performance. In this study, we tested the early prediction performance of Scaled-Dot Product Attention with conventional algorithms. We then investigated essential lectures or features related to students' learning activities. From the result, we found that Scaled-Dot Product Attention outperformed the conventional algorithms to identify at-risk students and found the important lectures and students' actions.

DOI： 10.1109/TALE54877.2022.00059

Scopus

Publisher：CEUR Workshop Proceedings  

Student Performance Prediction is essential for learning analysis of the students' learning behavior to discovering at-risk students for the early invention to support students. This study transforms the students' reading behavior into a two-dimensional matrix input based on each lecture material's reading behavior. The matrix input will be updated by accumulating the value for each week for performance prediction week by week. The multilayer perceptron neural network is employed to receive the matrix input and give feedback as a student's criteria consist of at-risk or no-risk students. This study considers the accuracy of a model considering between on contents information and weekly information. We also investigate the switching of learning materials' order, the feature importance of the reading operation on an event stream, and the difference in reading behavior between at-risk and no-risk students. These can help the instructors for an early invention to support at-risk students.

Scopus

Publishing type：Research paper (international conference proceedings)   Publisher：ACM  

Handwritten notes are one important component of students' learning process, which is used to record what they have learned in class or tease out knowledge after class for reflection and further strengthen the learning effect. It also helps a lot during review. We hope to divide handwritten notes (Japanese) into different parts, such as text, mathematical expressions, charts, etc., and quantify them to evaluate the condition of the notes and compare them among students. At the same time, data on students' learning behaviors in the course are collected through the online education platform, such as the use time of textbook and attendance, as well as the scores of the online quiz and course grade. In this paper, the analysis of the relationship between the segmentation results of handwritten notes and learning behavior are reported, as well as the research on automatic page segmentation based on deep learning.

DOI： 10.1145/3506860.3506915

Web of Science

Scopus

researchmap

Other Link： https://dl.acm.org/doi/pdf/10.1145/3506860.3506915

Publisher：CEUR Workshop Proceedings  

The emergence of digital textbook reading systems such as Bookroll, and their ability of recording reader interactions has opened the possibility of analyzing the students reading behaviors and characteristics. To date, several works have conducted compelling analyses characterizing the different types of students with the use of clustering ML models, while others have used supervised ML models to predict their academic performance. The main characteristic these models share is that internally they simplify the students' data into a latent representation to get an insight or make a prediction. Nevertheless, these representations are oversimplified, otherwise difficult to interpret. Accordingly, the present work explores the use of Variational Autoencoders to make more interpretable and complex latent representations. After a brief description of these models, we present and discuss the results of four explorative studies when using the LAK22 Data Challenge Workshop datasets. Our results show that the probabilistic latent representations generated by the proposed models preserve the student reading characteristics, allowing a better visual interpretation when using 3 dimensions. Also, they allow supervised regressive and classification models to have a more stable and less overfitted learning process, which also allows some of them to make better score predictions.

Scopus

Web of Science

Publisher：AVSS 2022 - 18th IEEE International Conference on Advanced Video and Signal-Based Surveillance  

Background subtraction networks outperform traditional hand-craft background subtraction methods. The main advantage of background subtraction networks is their ability to automatically learn background features for training scenes. When applying the trained network to new target scenes, adapting the network to the new scenes is crucial. However, few studies have focused on reusing multiple trained models for new target scenes. Considering background changes have several categories, such as illumination changes, a model trained for each background scene can work effectively for the target scene similar to the training scene. In this study, we propose a method to ensemble the module networks trained for each background scene. Experimental results show that the proposed method is significantly more accurate compared with the conventional methods in the target scene by tuning with only a few frames.

DOI： 10.1109/AVSS56176.2022.9959316

Web of Science

Scopus

Web of Science

Event date： 2022.12

Language：English  

Country：Japan  

Event date： 2022.10

Language：Japanese  

Country：Japan  

Event date： 2022.8

Language：English  

Country：Japan  

Event date： 2022.8

Language：English  

Country：Japan  

Event date： 2022.3

Language：English  

Event date： 2022.3

Language：English  

Event date： 2022.3

Language：English  

Event date： 2022.3

Language：English  

Event date： 2022.3

Language：English  

Event date： 2022.3

Language：English  

Event date： 2021.7

Language：English  

Event date： 2021.6

Language：English  

Event date： 2021.5

Language：Japanese  

Event date： 2021.5

Language：Japanese  

Event date： 2021.5

Language：Japanese  

Event date： 2021.4

Language：English  

Event date： 2021.1

Language：English  

Event date： 2020.12

Language：English  

Event date： 2020.12

Language：English  

Event date： 2020.12

Language：English  

Event date： 2020.11

Language：English  

Event date： 2020.8

Language：English  

Event date： 2020.7

Language：English  

Event date： 2020.5

Language：Japanese  

Event date： 2020.3

Language：English  

Event date： 2020.3

Language：English  

Event date： 2020.3

Language：English  

Event date： 2020.3

Language：English  

Event date： 2019.11

Language：English  

Event date： 2019.11

Language：English  

Event date： 2019.9

Language：English  

Event date： 2019.9

Language：English  

Event date： 2019.9

Language：English  

Event date： 2019.9

Language：English  

Event date： 2019.7

Language：English  

Event date： 2019.7

Language：English  

Event date： 2019.7

Language：English  

Event date： 2019.3

Language：English  

Event date： 2019.3

Language：English  

Event date： 2019.3

Language：English  

Event date： 2019.3

Language：English  

Event date： 2018.12

Language：English  

Event date： 2018.10

Language：English  

Event date： 2018.10

Language：English  

Event date： 2018.9

Language：English  

Event date： 2018.9

Language：English  

Event date： 2018.7

Language：English  

Event date： 2018.7

Language：English  

Event date： 2018.3

Language：English  

Event date： 2018.3

Language：English  

Event date： 2018.3

Language：English  

Event date： 2018.1

Language：English  

Event date： 2017.12

Language：English  

Event date： 2017.12

Language：English  

Event date： 2017.10

Language：English  

Event date： 2017.10

Language：English  

Event date： 2017.9

Language：English  

Event date： 2017.8

Language：Japanese  

Event date： 2017.8

Language：English  

Event date： 2017.7

Language：English  

Event date： 2017.7

Language：English  

Event date： 2017.5

Language：English  

Event date： 2017.3

Language：English  

Event date： 2017.3

Language：English  

Event date： 2016.10

Language：Japanese  

Event date： 2015.9

Language：Japanese  

Event date： 2015.9

Language：Japanese  

Event date： 2015.3

Language：English  

Event date： 2013.6

Language：Japanese  

Event date： 2011.10

Language：English  

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

Language：Japanese  

Language：Japanese  

Type：Competition, symposium, etc. 

Type：Competition, symposium, etc.