In recent years, early-warning systems have emerged at the K-12 level. However, these systems typically warn school districts about only some potential forms of risk, most commonly dropout or school violence. While some efforts have extended to predicting college enrollment or SAT score, a student can graduate from high school, attend college, achieve an acceptable SAT score, and still find that their earlier learning experiences shape their opportunities to choose a broad range of careers.

In 2017-2018, researchers at Worcester Polytechnic Institute and the University of Pennsylvania released an 11-year longitudinal data set spanning from middle school usage of a blended learning system for mathematics (ASSISTments) to the students’ eventual choice of career, focusing on whether experiences in mathematics class corresponded to eventually choosing a STEM (Science, Technology, Engineering, and Mathematics) job or graduate program after college. This data set was released as part of a competition affiliated with the NSF Northeast Big Data Hub and the Spoke on Big Data for Education. Over two hundred participants registered in the competition and 74 participants eventually submitted solutions.

In this workshop, we invite interested researchers to share their findings related to the competition’s publicly-released data sets, both in terms of data mining and education research, with the broader scientific community. Although we anticipate that many of the submissions will come from participants in the competition, this special issue is open to any researcher analyzing data from the competition’s data sets, which have remained available after the competition’s conclusion.

Relevant paper topics may include but are not limited to:

• Detecting and predicting future student participation in STEM careers
• Identifying the characteristics of students who will participate in STEM
• Identifying the attributes of students who are successful and/or engaged in STEM in middle school but who will lose interest in later years
• How the findings of this competition may contribute to the design of systems that increase interest in studying and working in STEM fields
• Long-term and longitudinal interventions related to this competition
• Longitudinal development and change in students from middle school to post-college
• Discoveries about machine learning, time series modeling, feature extraction and selection, and student modeling produced through participating in the competition
• Higher-level papers on the contributions of competitions such as this one to educational research


Review Process
As stipulated by JEDM reviewing guidelines, each submission will be peer-reviewed by three experts in the field. Reviewers will be selected from the JEDM Editorial Board, participants in the competition, and other researchers selected by the special issue organizers for their relevant expertise.

Submission Guidelines
We invite submissions of any length. Please see the JEDM submission guidelines. All submissions should be made through the JEDM article submission system.

Please note in your cover letter that your article is intended for this special issue.

http://jedm.educationaldatamining.org/
http://jedm.educationaldatamining.org/index.php/JEDM/about/submissions

For authors who have published an initial version of the results at the competition workshop at EDM2018, please note that we only can consider submissions that are significantly extended with at least 33% new material compared to the workshop proceedings. In any case, this special issue is a public/open call for papers and researchers do not need to have participated in the workshop to submit to (and be accepted by) this special issue.

Deadlines
Please submit all articles for consideration in this special issue by December 31, 2018.

Contact
Please direct any questions relevant to this special issue to assistmentscomp2018@wpi.edu