TUTORIAL 1

Title:  Learning Analytics Hands-On Tutorial (Half-a-day )

Coordinator:

• Professor Alexandra I. Cristea, Durham University (Bio on webpage: Professor AI Cristea – Durham University)
• Experience: currently running a 4^th year Learning Analytics Module; have given LA keynotes and tutorials in the past.

Team: a team of PhD students and guest researchers in Learning Analytics, demonstrating

Target Audience:

• Young Researchers
• Researchers wanting to move from more traditional ITS approaches towards big data analytics for learning
• Teachers, practitioners, education decision-makers
• Min numbers to run: 5 participants;
• Max numbers able to run: 50 participants;

Abstract:

• Motivation:

  • Data is becoming ubiquitous, and data analytics has almost swallowed up the term ‘AI’ in the understanding of the press and mass-media. Its applications are everywhere – in industry, governance, and, more recently, in education – the latter called Learning Analytics. ‘Old fashioned’ ‘top-down’ approaches, such as Intelligent Tutoring Systems (ITS) and Artificial Intelligence in Education (AIED), based on solid pedagogical foundations, have almost been replaced in many ways by the new ‘bottom-up’ approach of letting data guide us. Whilst I believe that the truth is somewhere in the middle, in this tutorial I wish to address what this new hype is about, some of its main methods, as well as allow participants some hands-on experimentation.

• Aims (Expected Outcomes):

  • The aims of this tutorial is to give participants a fundamental understanding of some of the core approaches and problem-solving principles for Learning Analytics (LA) and the role of LA in current and future learning settings and environments.

• Content:

  • Statistical Learning Analytics and visualisation: data pre-processing; methods for tackling learning analytics based on statistical approaches; the types of LA that can be done with such approaches – e.g. descriptive and beyond. Visualisation of LA data for different stakeholders – e.g. learner, teacher, administrator, etc.

  • Ethics of Learner Data Usage: discussions on ethical considerations of using learner data, starting from societal view, laws involved, (common) practice, future practice. Algorithmic perspectives, such as (expanded) sensitivity analysis.

  • Machine Learning-based Learning Analytics: shallow and deep Machine Learning methods for LA; numerical versus textual data analytical methods for LA; combined methods; sentiment analysis for LA; the types of LA that can be done with such approaches – e.g. descriptive, diagnostic, predictive, prescriptive.

Format:

• Talks interspersed with hands-on sessions (Note: for the hands-on sessions, Google Colab will be used, so participants will be asked to ensure to have a Google account in advance to be able to participate hands-on; otherwise, they can see the demonstrations projected by the teaching team)

Tutorial  1

TUTORIAL  2

Title: Data Science for learning process management (half-a-day)

Coordinator:

Filippo Sciarrone, Roma TRE University – Department of Engineering
Via della Vasca navale 79, 00146 Rome, Itay

Theme and goals:

Nowadays, all learning platforms are producing large amounts of data, i.e., big data. The scientific community that studies learning processes has the opportunity to have several available tools to extract useful information from data to improve the learning process itelf, in its broadest sense. Consequently, all the stakeholders involved have the opportunity to improve their contribution. The Internet provides business intelligence and data science platforms, aimed at processing and studying large amounts of data, regardless of the application domain which generated them, such as Weka, R, Knime, etc. etc. This tutorial aims to show the power of one of these tools, the Knime platform, free and available at www.knime.org, for applications to the learning domain of huge students’ communities such as Massive Online Open Courses. The proposal is aimed at the ITS community as a directly involved community in this field of research and especially in the field of educational data mining and learning and teaching analytics.

Fields of research strongly involved: Technology Enhanced Learning, Learning Analytics, Educational Data Mining

The tutorial will cover:

– Educational Data Mining: the most used techniques for studying learning processes

– The KNIME platform: workflows and functionalities at a basic level

– Simple case studies

Target audience:

Participants must have two important requirements: have the KNIME platform installed on their laptop or desktop and have a minimum of background in machine learning techniques, better if applied to learning processes.

Activities planned.  

• • 09:00-11:00
    • Introduction to the The Knime Platform
    • Machine Learning useful algorithms in Knime for education
    • Workflow Building
    • Data management
  • 11:00-13:00
    • Case study 1: MOOCS data
    • Case study 2: Moodle data

Expected outcomes.  

Through this tutorial, participants will acquire some basic skills and knowledge to be able to design a data management workflow applied to the educational field, by using the Knime business intelligence platform.

Video/audio facilities and other equipment needed

Participants need  some data sets for the case studies analysis, posted by myself before starting

Tutorial  2

Last day to register for a tutorial: Monday, May 24, 2021