Research Team Led by Professor Shin Hyo-jung Has Papers Accepted at AIED 2026, a Leading International Conference on Educational Artificial Intelligence
A research team led by Professor Shin Hyo-jung—who serves as the Head Professor of both the Interdisciplinary Major in Education and Culture at the College of Humanities and the Interdisciplinary AI Behavioral Studies Program at the Graduate School—has had two papers simultaneously accepted for publication at the 27th International Conference on Artificial Intelligence in Education (AIED 2026), a globally renowned conference in the field of educational artificial intelligence.
This achievement is particularly noteworthy given the unprecedented competition: a record 1,241 papers were submitted—a 35% increase from the previous year (AIED 2025) and a 200% increase from 2024 (AIED 2024). In particular, both papers were accepted as “Full Papers” for the Main Conference following a rigorous review process in which only 16.7% of all submitted papers were accepted, demonstrating high recognition of the research’s academic value.
Above all, the significance of these published studies is heightened as they are the outcomes of Professor Shin Hyo-jung’s large-scale global research project and the undergraduate Capstone Design course.
● First Paper
Professor Shin Hyo-jung served as the first author and corresponding author, and Kim Eui-gyeom (67th Cohort, Graduate School of Education, majoring in AI Integration in Educational Design and Management; BA in American Culture, '16) participating as a co-author. Titled “How Linguistic Diversity Impacts Multilingual Automated Scoring in Large-Scale Assessments,” the paper analyzed the mechanisms of linguistic diversity as a key factor influencing the accuracy of automated scoring in large-scale international comparative tests.
The research team analyzed large-scale assessment data from the 2016 ePIRLS (The Progress in International Reading Literacy Study), which was administered in 14 languages and involved approximately 70,000 students from 15 countries. The study is significant in that it did not view the performance of automated scoring on students’ constructed responses as merely a problem with the scoring model, but rather quantitatively identified linguistic diversity as a fundamental factor.
This study provides a foundation for establishing practical operational guidelines to determine the extent to which automation is feasible and when human scoring is necessary when introducing AI-based automated scoring in future large-scale international comparative tests and national-level academic achievement assessments. Furthermore, this study is expected to contribute to the development of a more valid, reliable, and explainable educational assessment system through human-AI collaborative assessment.
▶ Paper Title: How Linguistic Diversity Impacts Multilingual Automated Scoring in Large-Scale Assessments
▶Authors: Shin Hyo-jung (First Author and Corresponding Author), Nico Andersen (Co-author, DIPF), Kim Eui-gyeom (Co-author, Sogang University), Andrea
Horbach (IPN - Leibniz Institute for Science and Mathematics Education), and Fabian Zehner (DIPF)
▶ AIED 2026 Website: https://aied-conference.org/2026
● Second Paper
The paper titled “I’m Not Stuck – I’m Learning: Operationalizing Self-Efficacy Trajectories in Large-Scale AI Learning Environments,” in which Professor Shin Hyo-jung served as corresponding author, is the outcome of a capstone project conducted in her course “Understanding Big Data and Its Educational Applications (Capstone Design) - EDU3048” during the fall semester of the 2025 academic year. This project was a collaborative achievement with Sogang University undergraduates Hong Bu-won (Life Science, ’20), Shim Eun (Computer Science and Engineering, ’22), and Ko Hee-soo (English Literature and Linguistics, ’22), and is significant in that it demonstrates the connection between the university’s undergraduate courses and academic achievements.
This study utilized the EdNet dataset, which contains learning data from approximately 780,000 learners, to overcome the limitations of self-reported self-efficacy surveys and propose metrics measured through learning log data.
The research team found that while the sheer volume of problem-solving (Practice intensity) is closely related to final achievement, actual improvement in learning ability (Learning improvement) is determined by the process of finding the correct answer on one’s own immediately after making a mistake, as well as the ability to self-regulate one’s learning rhythm. Furthermore, they discovered that irregularities in learning routines predict dropout more accurately than a decline in grades, and proposed a new learning management system capable of monitoring learners’ psychological states and routine disruptions in real time. This research is expected to serve as a key foundation for building intelligent tutoring systems that prevent learners from dropping out and support continuous growth by providing tailored interventions at appropriate points during the learning process.
▶Paper Title: I’m Not Stuck – I’m Learning: Operationalizing Self-Efficacy Trajectories in Large-Scale AI Learning EnvironmentsPaper Title: I’m Not Stuck – I’m Learning: Operationalizing Self-Efficacy Trajectories in Large-Scale AI Learning Environments
▶Authors: Hong Bu-won (First Author), Shim Eun (Co-author), Ko Hee-soo (Co-author), and Professor Shin Hyo-jung (Corresponding Author)
▶ AIED 2026 Website: https://aied-conference.org/2026
This achievement can be attributed to Sogang University’s unique innovative educational model, which combines Professor Shin’s research expertise with the students’ passion for inquiry. She emphasized that this achievement is “a valuable example of how learning in the classroom leads to practical research that solves real-world problems,” adding, “We will continue to explore the mutual cooperation between AI and humans through Capstone Design and Inquiry Community courses with Sogang University undergraduates, and pursue research to create a better educational environment.”
Meanwhile, AIED 2026, including workshops, is scheduled to be held at COEX in Seoul from June 27 to July 3, 2026.
[SEO Keyword]
AIED 2026, Automated Scoring Accuracy, Intelligent Tutoring Systems
This achievement is particularly noteworthy given the unprecedented competition: a record 1,241 papers were submitted—a 35% increase from the previous year (AIED 2025) and a 200% increase from 2024 (AIED 2024). In particular, both papers were accepted as “Full Papers” for the Main Conference following a rigorous review process in which only 16.7% of all submitted papers were accepted, demonstrating high recognition of the research’s academic value.
Above all, the significance of these published studies is heightened as they are the outcomes of Professor Shin Hyo-jung’s large-scale global research project and the undergraduate Capstone Design course.
● First Paper
Professor Shin Hyo-jung served as the first author and corresponding author, and Kim Eui-gyeom (67th Cohort, Graduate School of Education, majoring in AI Integration in Educational Design and Management; BA in American Culture, '16) participating as a co-author. Titled “How Linguistic Diversity Impacts Multilingual Automated Scoring in Large-Scale Assessments,” the paper analyzed the mechanisms of linguistic diversity as a key factor influencing the accuracy of automated scoring in large-scale international comparative tests.
The research team analyzed large-scale assessment data from the 2016 ePIRLS (The Progress in International Reading Literacy Study), which was administered in 14 languages and involved approximately 70,000 students from 15 countries. The study is significant in that it did not view the performance of automated scoring on students’ constructed responses as merely a problem with the scoring model, but rather quantitatively identified linguistic diversity as a fundamental factor.
This study provides a foundation for establishing practical operational guidelines to determine the extent to which automation is feasible and when human scoring is necessary when introducing AI-based automated scoring in future large-scale international comparative tests and national-level academic achievement assessments. Furthermore, this study is expected to contribute to the development of a more valid, reliable, and explainable educational assessment system through human-AI collaborative assessment.
▶ Paper Title: How Linguistic Diversity Impacts Multilingual Automated Scoring in Large-Scale Assessments
▶Authors: Shin Hyo-jung (First Author and Corresponding Author), Nico Andersen (Co-author, DIPF), Kim Eui-gyeom (Co-author, Sogang University), Andrea
Horbach (IPN - Leibniz Institute for Science and Mathematics Education), and Fabian Zehner (DIPF)
▶ AIED 2026 Website: https://aied-conference.org/2026
● Second Paper
The paper titled “I’m Not Stuck – I’m Learning: Operationalizing Self-Efficacy Trajectories in Large-Scale AI Learning Environments,” in which Professor Shin Hyo-jung served as corresponding author, is the outcome of a capstone project conducted in her course “Understanding Big Data and Its Educational Applications (Capstone Design) - EDU3048” during the fall semester of the 2025 academic year. This project was a collaborative achievement with Sogang University undergraduates Hong Bu-won (Life Science, ’20), Shim Eun (Computer Science and Engineering, ’22), and Ko Hee-soo (English Literature and Linguistics, ’22), and is significant in that it demonstrates the connection between the university’s undergraduate courses and academic achievements.
This study utilized the EdNet dataset, which contains learning data from approximately 780,000 learners, to overcome the limitations of self-reported self-efficacy surveys and propose metrics measured through learning log data.
The research team found that while the sheer volume of problem-solving (Practice intensity) is closely related to final achievement, actual improvement in learning ability (Learning improvement) is determined by the process of finding the correct answer on one’s own immediately after making a mistake, as well as the ability to self-regulate one’s learning rhythm. Furthermore, they discovered that irregularities in learning routines predict dropout more accurately than a decline in grades, and proposed a new learning management system capable of monitoring learners’ psychological states and routine disruptions in real time. This research is expected to serve as a key foundation for building intelligent tutoring systems that prevent learners from dropping out and support continuous growth by providing tailored interventions at appropriate points during the learning process.
▶Paper Title: I’m Not Stuck – I’m Learning: Operationalizing Self-Efficacy Trajectories in Large-Scale AI Learning EnvironmentsPaper Title: I’m Not Stuck – I’m Learning: Operationalizing Self-Efficacy Trajectories in Large-Scale AI Learning Environments
▶Authors: Hong Bu-won (First Author), Shim Eun (Co-author), Ko Hee-soo (Co-author), and Professor Shin Hyo-jung (Corresponding Author)
▶ AIED 2026 Website: https://aied-conference.org/2026
This achievement can be attributed to Sogang University’s unique innovative educational model, which combines Professor Shin’s research expertise with the students’ passion for inquiry. She emphasized that this achievement is “a valuable example of how learning in the classroom leads to practical research that solves real-world problems,” adding, “We will continue to explore the mutual cooperation between AI and humans through Capstone Design and Inquiry Community courses with Sogang University undergraduates, and pursue research to create a better educational environment.”
Meanwhile, AIED 2026, including workshops, is scheduled to be held at COEX in Seoul from June 27 to July 3, 2026.
[SEO Keyword]
AIED 2026, Automated Scoring Accuracy, Intelligent Tutoring Systems