Learning-Oriented Question Recommendation Using Bloom’s Learning Taxonomy and Variable Length Hidden Markov Models - Transactions on Large-Scale Data-and Knowledge-Centered Systems - page 29

Learning-Oriented Question Recommendation
Using Bloom's Learning Taxonomy
and Variable Length Hidden Markov Models

B

Hilda Kosorus (

) and Josef Kung

Institute of Application Oriented Knowledge Processing,
Johannes Kepler University, Linz, Austria
{ hkosorus,jkueng } @faw.jku.at
http://www.faw.jku.at

Abstract. The information overload in the past two decades has enabled
question-answering (QA) systems to accumulate large amounts of tex-
tual fragments that reflect human knowledge. Therefore, such systems
have become not just a source for information retrieval, but also a means
towards a unique learning experience. Recently developed recommenda-
tion techniques for search engine queries try to leverage the order in which
users navigate through them. Although a similar approach might improve
the learning experience with QA systems, questions would still be consid-
ered as abstract objects, without any content or meaning. In this paper,
a new learning-oriented technique is defined that exploits not only the
user's history log, but also two important question attributes that reflect
its content and purpose: the topic and the learning objective. In order to
do this, a domain-specific topic-taxonomy and Bloom's learning frame-
work is employed, whereas for modeling the order in which questions are
selected, variable length Markov chains (VLMC) are used. Results show
that the learning-oriented recommender can provide more useful, mean-
ingful recommendations for a better learning experience than other pre-
dictive models.

· ·

Keywords: Recommender systems Question-answering systems

· · ·

Learning taxonomy Topic taxonomy Collaborative filtering Vari-

able length markov chain

1 Introduction

In the past two decades, education has become more and more subject to per-
sonalization and automation. The success of recommender systems [ 1 ] has moti-
vated research on deploying such techniques also in educational environments to
facilitate access to a wide spectrum of information [ 16 ].
One of the consequences of information overload is the rise of question answer-
ing (QA) systems. Over time, QA systems have gathered a large amount of tex-
tual fragments - reflections of human knowledge - from a variety of domains
c