Current students

The term Learning Analytics (LA) indicates the field of research that deals with measuring, collecting, analysing and reporting data regarding the educational context, i.e. regarding students, teachers, school managers and the school environment in which they study or work. The main objective of the LA is to improve the quality of learning and the educational environment. [8]
One of the most important applications of learning analytics is the prediction of students' academic performance in order to improve the learning process both for students at risk and for those who are already sufficient or excellent. Furthermore, not only accurate but also early in time predictions are a valuable resource for improving the quality of learning, there will be more time to prevent the failure of at-risk students and motivate excellent students with personalized additional activities [3]. An open question is the search for the right balance between accuracy and timeliness, this is the main question in the context of early warning systems (EWS). To answer this question, recently, some studies such as [3] [5] [8] [10] are applying the idea of ​​updating the prediction estimates with periodicity.
The Internet and new technologies have made numerous changes to the learning process. In recent years, completely online courses and blended courses were established and the technology was also used to support face-to-face activities. The Coronavirus (COVID-19) pandemic of recent years has forced atotal closure of schools and universities, banning face-to-face courses and speeding up the use of online learning. ([4] [6] [9]) The use of digital technologies in the educational context has changed the way of teaching and learning and has allowed the collection of information and data from online learning environments such ad Learning Management System. Krishnan et al. ([7]) define Learning Management System (LMS) as “a software application designed to manage, monitor, report and provide courses and teaching materials”. It is important to understand what extend this data combined with administrative data can improve the predictions of students' academic performance.
Recently, many studies ([3] [4] [5] [8] [10]) aimed at predicting student performance have demonstrated the effectiveness of using Machine Learning methods. ML is a branch of Artificial Intelligence (AI) that deal with data analysis and give computers the ability to learn from data the interactions between the response and the predictors rather than assume any functional form to estimate the parameters from data.
The first contribution of this work fits into the contexts just presented. With the use of administrative data combined with data from LMS it aims at researching the best ML algorithm to use and investigating what is best time to find the balance between accuracy and timeliness. The data used will be those regarding students of the Politecnico di Milano whose digital support device for teaching is the WeBeep platform; the proposed algorithms will be supervised ML algorithms updated on a weekly basis. In particular, this work aims to evaluate the extent to which data regarding the use of LMS can support predictions in blended or face to face courses.
After having classified the students, an intervention plan based on social comparison is proposed and evaluated. According to the social comparison theory of Festinger ([2]), men tend to value themselves based on comparison with peers. As demonstrated by Davis et al. ([1]) "personalized social-comparison feedback increases course completion rates ". And it is precisely inspired by Davis et al. ([1]) 's work that this work proposes the creation of a dashboard to show a student how much the platform has been used by colleagues who have passed the course in previous years, comparing with how much the student is using it. The dashboard will be shown to students on a weekly basis (in accordance with the evaluation periodicity of the predictive model); the impact that this social comparison feedback has on the use of the platforms and consequently on the prediction of performance will be assessed.
The latest contribution of this research aims to retrace the same analyses in a different context, namely high schools, in which, up to my knowledge, there are still no studies like this one.    ;

;
1. Dan Davis, Ioana Jivet, René F. Kizilcec, Guanliang Chen, Claudia Hauff, and Geert-Jan Houben. 2017. Follow the successful crowd: raising MOOC completion rates through social comparison at scale. In Proceedings of the Seventh International Learning Analytics & Knowledge Conference (LAK '17). Association for Computing Machinery, New York, NY, USA, 454–463. DOI: https://doi.org/10.1145/3027385.3027411
;
2. Festinger L. A Theory of Social Comparison Processes. Human Relations. 1954;7(2):117-140. doi: 1177/001872675400700202
;
3. Moises Riestra-González, Maria del Puerto Paule-Ruíz, Francisco Ortin, Massive LMS log data analysis for the early prediction of course-agnostic student performance, Computers & Education, Volume 163, 2021, 104108, ISSN 0360-1315, https://doi.org/10.1016/j.compedu.2020.104108.
;
4. Saleem, F.; Ullah, Z.; Fakieh, B.; Kateb, F. Intelligent Decision Support System for Predicting Student’s E-Learning Performance Using Ensemble Machine Learning. Mathematics 2021, 9, 2078. https://doi.org/10.3390/math9172078
;
5. Evandro B. Costa, Baldoino Fonseca, Marcelo Almeida Santana, Fabrísia Ferreira de Araújo, Joilson Rego, Evaluating the effectiveness of educational data mining techniques for early prediction of students' academic failure in introductory programming courses, Computers in Human Behavior, Volume 73, 2017, Pages 247-256, ISSN 0747-5632, https://doi.org/10.1016/j.chb.2017.01.047.
;
6. Najib Ali Mozahem, 2020. "Using Learning Management System Activity Data to Predict Student Performance in Face-to-Face Courses," International Journal of Mobile and Blended Learning (IJMBL), IGI Global, vol. 12(3), pages 20-31, July.
;
7. Krishnan, R.; T, S.; Nair, S.; Saamuel, B.S.; Iwendi, C.; Biamba, C.; Ibeke, E. Smart Analysis of Learners Performance Using Learning Analytics for Improving Academic Progression: A Case Study Model. Sustainability 2022, 14, 3378. https://doi.org/10.3390/su14063378.
;
8. Nikola Tomasevic, Nikola Gvozdenovic, Sanja Vranes, An overview and comparison of supervised data mining techniques for student exam performance prediction, Computers & Education, Volume 143, 2020, 103676, ISSN 0360-1315, https://doi.org/10.1016/j.compedu.2019.103676.
;
9. Benjamín, Maraza-Quispe, Enrique Damián Valderrama-Chauca, Lenin Henry Cari-Mogrovejo, Jorge Milton Apaza-Huanca, and Jaime Sanchez-Ilabaca. “A Predictive Model Implemented in KNIME Based on Learning Analytics for Timely Decision Making in Virtual Learning Environments.” International Journal of Information and Education Technology 12, no. 2 (2022): 91–99. doi: 10.18178/ijiet.2022.12.2.1591.
;
10. Park, Hee & Yoo, Seong. (2021). Early Dropout Prediction in Online Learning of University using Machine Learning. JOIV : International Journal on Informatics Visualization. 5. 347. 10.30630/joiv.5.4.732.
;