Educational data mining

Summary

More about Associate Professor Kalina Yacef Stay in touch

This project aims at mining the huge amount of electronic data collected by educational systems to discover useful information for teachers, learners and education researchers. Education being a novel application domain for Data Mining, a large part of this project is focused on creating or adapting suitable mining algorithms.

Supervisor

Associate Professor Kalina Yacef.

Research location

Computer Science

Program type

N/A

Synopsis

Web-based educational systems collect tremendous amount of electronic data, ranging from simple histories of students' interactions with the system to detailed traces about their reasoning. However, less attention has been given to handling the large quantities of data collected from the students' interactions and extracting pedagogically useful information from it. Such systems give teachers and learning researchers access to an extensive source of electronic data about students' learning, data which is currently under-exploited. Data mining techniques have the potential to remedy this situation.

Data mining encompasses a range of techniques and algorithms for discovering interesting patterns hidden in large data sets such as association rules, classification, cluster analysis as well as statistical analysis and database query.

In this project, the goals are:

• To identify, adapt or create new data mining methods that are suited for turning learners' performance data into information of relevance to teachers, instructional designers, and learning researchers.
• To define how to "massage" the student data so that we can extract interesting patterns.
• To automate/facilitate some of the algorithm selection and pre-processing features.
• To find suitable ways to present the results to users.
• To exploit the patterns found to improve adaptation of teaching systems

Want to find out more?

Opportunity ID

The opportunity ID for this research opportunity is 508