Παρακαλώ χρησιμοποιήστε αυτό το αναγνωριστικό για να παραπέμψετε ή να δημιουργήσετε σύνδεσμο προς αυτό το τεκμήριο:
https://hdl.handle.net/123456789/1870| Τύπος: | Πρακτικά συνεδρίου |
| Τίτλος: | Α comparison of two text-based feature sets for predicting student performance: a initial exploration |
| Συγγραφέας: | [EL] Καρασαββίδης, Ηλίας[EN] Karasavvidis, Ilias [EL] Παπαδήμας, Χαράλαμπος[EN] Papadimas, Charalampos [EL] Ραγάζου, Βασιλική[EN] Ragazou, Vasiliki |
| Ημερομηνία: | Νοε-2021 |
| Περίληψη: | Research interest in Learning Analytics has skyrocketed in the course of the past decade. While former research has considered several features for predicting student performance, student-generated texts have not been explored as a source. The purpose of this study is to explore text as a feature for predicting student performance. More specifically our goal in this work is to compare the predictive power of raw text compared to engineered text features. Forty two student teachers from an educational department at a Greek university participated in the study. The participants watched a series of six video lectures and were required to write a short summary for each. Their understanding of each video lecture was measured using a quiz. Based on their median performance on this quiz, the students were split into two classes: low and high performance. A large number of machine learning classification algorithms (e.g. Linear Regression, Support Vector Classifier, Naïve Bayes, K-means, Random Forest) were employed to determine which text feature set (raw text vs engineered features) provides higher classification accuracy. The results show that, while the information in the raw text is impressive, the engineered features may help go the extra mile in terms of classification accuracy. |
| Γλώσσα: | Αγγλικά |
| Τόπος δημοσίευσης: | Online Conference |
| Σελίδες: | 9 |
| DOI: | 10.21125/iceri.2021.1101 |
| ISBN: | 978-84-09-34549-6 |
| ISSN: | 2340-1095 |
| Θεματική κατηγορία: | [EL] Κοινωνικές επιστήμες, άλλοι τομείς[EN] Social sciences, miscellaneous |
| Λέξεις-κλειδιά: | machine learning; Supervised learning; unstructured text; feature engineering; engineered text features; Text mining; text similarity; natural language processing |
| Κάτοχος πνευματικών δικαιωμάτων: | © The Author(s) 2021 |
| Ηλεκτρονική διεύθυνση του τεκμηρίου στον εκδότη: | https://library.iated.org/view/KARASAVVIDIS2021COM |
| Ηλεκτρονική διεύθυνση περιοδικού: | https://iated.org/archive/iceri2021 |
| Τίτλος πηγής δημοσίευσης: | ICERI Proceedings |
| Σελίδες τεκμηρίου (στην πηγή): | 4806-4814 |
| Όνομα εκδήλωσης: | ICERI2021, the 14th International Conference of Education, Research and Innovation |
| Τοποθεσία εκδήλωσης: | On-line conference |
| Ημ/νία έναρξης εκδήλωσης: | 08/11/2021 |
| Ημ/νία λήξης εκδήλωσης: | 09/11/2021 |
| Σημειώσεις: | This research is co-financed by Greece and the European Union (European Social Fund- ESF) through the Operational Programme "Human Resources Development, Education and Lifelong Learning 2014-2020" in the context of the project “Planning, Development and Deployment of an Intelligent Feedback System Using Supervised and Unsupervised Machine Learning Methods” (MIS 5048955). |
| Εμφανίζεται στις συλλογές: | Ερευνητικές ομάδες |
Αρχεία σε αυτό το τεκμήριο:
Το πλήρες κείμενο αυτού του τεκμηρίου δεν διατίθεται προς το παρόν από το αποθετήριο