Conference Session

Topics in Computing

Collection

2019 ASEE Annual Conference & Exposition

Authors

Naeem Seliya, Ohio Northern University; Heath Joseph LeBlanc, Ohio Northern University; J. Blake Hylton, Ohio Northern University; Ziad Youssfi, Ohio Northern University; Matthew Schweinefuss

Tagged Topics

Diversity

Tagged Divisions

Computing and Information Technology

filters provide relatively less similarity ingeneral. Among all five feature selection methods examined, GainRatio is determined as the bestapproach for our case study, because it identifies words relevant to the subject that highlycorrelate to a particular level (class) of Propensity for Exploration even if they are sparselyrepresented in the dataset. These words can be viewed as the diamonds in the rough thatdistinguish the question. We note that like most data mining based studies, the case study resultsare determined on the underlying dataset and the algorithms investigated. Our proposedapproach, however, can be applied to other curiosity exercise datasets as well, and provide therelevant experts a better insight into the student data.The