Question Generation from Concept Maps

In this paper we present a question generation approach suitable for tutorial dialogues. The
approach is based on previous psychological theories that hypothesize questions are generated from
a knowledge representation modeled as a concept map. Our model semi-automatically extracts
concept maps from a textbook and uses them to generate questions. The purpose of the study is
to generate and evaluate pedagogically-appropriate questions at varying levels of speci city across
one or more sentences. The evaluation metrics include scales from the Question Generation Shared
Task and Evaluation Challenge and a new scale speci c to the pedagogical nature of questions in
1. Introduction
A large body of research exists on question-related phenomena. Much of this research derives from
the tradition of question answering rather than question asking. Whereas question answering has
received extensive attention in computer science for several decades, with increasing interest over
the last 10 years (Voorhees and Dang 2005, Winograd 1972), question asking has received attention
primarily in educational/psychological circles (Beck et al. 1997, Bransford et al. 1991, Brown 1988,
Collins 1988, Dillon 1988, Edelson et al. 1999, Palinscar and Brown 1984, Piaget 1952, Pressley and
Forrest-Pressley 1985, Scardamalia and Bereiter 1985, Schank 1999, Zimmerman 1989) until the
recent surge of interest in computational accounts (Rus and Graesser 2009).
The distinction between question answering and question asking in some ways parallels the dis-
tinction between natural language understanding and natural language generation. Natural language
understanding takes a piece of text and maps it to one of many possible representations, while nat-
ural language generation maps from many possible representations to one piece of text … to download

About these ads