@article{cui_zhou_liu_jia_song_gehringer_2024, title={Utilizing the Constrained K-Means Algorithm and Pre-Class GitHub Contribution Statistics for Forming Student Teams}, url={https://doi.org/10.1145/3649217.3653634}, DOI={10.1145/3649217.3653634}, journal={PROCEEDINGS OF THE 2024 CONFERENCE INNOVATION AND TECHNOLOGY IN COMPUTER SCIENCE EDUCATION, VOL 1, ITICSE 2024}, author={Cui, Jialin and Zhou, Fangtong and Liu, Chengyuan and Jia, Qinjin and Song, Yang and Gehringer, Edward}, year={2024}, pages={569–575} }
 @article{cui_li_lou_liu_xiao_jia_gehringer_zhang_2022, title={Can Pre-class GitHub Contributions Predict Success by Student Teams?}, DOI={10.1109/ICSE-SEET55299.2022.9794264}, abstractNote={Over one million teachers, students, and schools around the world use GitHub to reach their learning goals. GitHub promotes team-work, and group or team projects are a necessary element of the software-engineering curriculum. Past studies on GitHub have explored how to integrate GitHub into teaching and how to mine information from GitHub to help students. To our knowledge, we are the first to study the previous contributions of students to GitHub in order to characterize student teams that perform well on team projects, compared to student teams that did not perform so well. We identify factors such as the number of public commits, number of repositories, and size of repositories in certain languages that are associated with the quality of team projects. We discuss the implications of this from the point of view of an educator and a student.}, journal={2022 ACM/IEEE 44TH INTERNATIONAL CONFERENCE ON SOFTWARE ENGINEERING: SOFTWARE ENGINEERING EDUCATION AND TRAINING (ICSE-SEET 2022)}, author={Cui, Jialin and Li, Ruochi and Lou, Kaida and Liu, Chengyuan and Xiao, Yunkai and Jia, Qinjin and Gehringer, Edward and Zhang, Runqiu}, year={2022}, pages={40–49} }
 @inproceedings{xiao_wang_sun_li_song_cui_jia_liu_gehringer_2022, title={Modeling review helpfulness with augmented transformer neural networks}, ISSN={["2325-6516"]}, DOI={10.1109/ICSC52841.2022.00019}, abstractNote={The past two years have witnessed a dramatic change in the delivery of education, as most providers have pivoted to remote online learning. With the enrollment of some MOOCs platforms, for example, Coursera, going up by 444% between mid-March and mid-September 202011Coursera 2020 Impact Report: https://about.coursera.org/press/wpcontent/uploads/2020/09/Coursera-Impact-Report-2020.pdf, practical teaching on a large scale attracted significant attention from the public. For online educators, this was manifested as a significant increase in assessment workload. MOOC instructors have long used peer assessment to evaluate work submitted by online learners. Prior research has shown that when a student believes feedback is helpful, then the suggestion given by that feedback is more likely to be implemented. Researchers have studied machine learning models for detecting problem statements or suggestions in feedback. However, these models do not work as well in detecting secondary features like helpfulness. This paper introduces a new augmented model for detecting helpfulness and tests it against the original model that uses raw text, yielding a 7% increase in performance measured in F1 score.}, booktitle={2022 IEEE 16th International Conference on Semantic Computing (ICSC)}, author={Xiao, Yunkai and Wang, Tianle and Sun, Xinhua and Li, Licong and Song, Yang and Cui, Jialin and Jia, Qinjin and Liu, Chengyuan and Gehringer, Edward}, year={2022}, pages={83–90} }