Theses in Progress

Machine Learning for Asking the Right Questions

March 01, 2019 – ,

Motivation

Did you ever wonder why learning something feels, in large parts, like it did 50 years ago? The goal of the Knowledge & Educational Technologies research group is to change that and to explore new ways of technology-enhanced learning.

We investigate ways to support learners and to make their learning visible and traceable, a research field called learning analytics. One important challenge in learning analytics is to figure out what learners already know about a certain subject. Having data about the known concepts of a learner would e.g. allow better content recommendation or more accurate dropout prediction. Thus, the goal of the advertised master thesis is to explore the capabilities of automatic question generation to simply ask learners about what they know
automatically.


Automatic question generation is an activate field of research that combines natural language processing and machine learning approachesto articulate natural language questions for a given answer and a given context.

 

Task

Create and evaluate an automatic question generator for educational purposes, for that:

  1. Review of the related work with special focus on

    • which approaches are most promising (template-based, transformation-based, machine learning based)?
    • what type of questions can be generated?

  2. Design and implementation of an automatic question generator
  3. Evaluation: either empirically or comparatively based on related work metrics

Requirements

  • Interest in Educational Technologies
  • Either knowledge of Natural Language Processing or knowledge of Machine Learning

 

 

download corresponding tendering

Keywords: Educational Technology, Machine Learning, Natural Language Processing

Research Area(s):

Tutor: Steuer, Christoph Rensing

Student: Viraj Kulkarni

Theses in Progress